University of Miami
Scholarly Repository
Open Access Teses Electronic Teses and Dissertations
2012-05-08
British Petroleum's Use of Image Restoration
Strategy on Social Media and Response Afer the
2010 Gulf Oil Spill
Katherine Rogers
University of Miami, [email protected]
Follow this and additional works at: htp://scholarlyrepository.miami.edu/oa_theses
Tis Open access is brought to you for free and open access by the Electronic Teses and Dissertations at Scholarly Repository. It has been accepted for
inclusion in Open Access Teses by an authorized administrator of Scholarly Repository. For more information, please contact
[email protected].
Recommended Citation
Rogers, Katherine, "British Petroleum's Use of Image Restoration Strategy on Social Media and Response Afer the 2010 Gulf Oil
Spill" (2012). Open Access Teses. Paper 334.
UNIVERSITY OF MIAMI
BRITISH PETROLEUM’S USE OF IMAGE RESTORATION STRATEGY ON
SOCIAL MEDIA AND RESPONSE AFTER THE 2010 GULF OIL SPILL
By
Katherine Rogers
A THESIS
Submitted to the Faculty
of the University of Miami
in partial fulfillment of the requirements for
the degree of Master of Arts
Coral Gables, Florida
May 2012
©2012
Katherine Rogers
All Rights Reserved
UNIVERSITY OF MIAMI
A thesis submitted in partial fulfillment of
the requirements for the degree of
Master of Arts
BRITISH PETROLEUM’S USE OF IMAGE RESTORATION STRATEGY ON
SOCIAL MEDIA AND RESPONSE AFTER THE 2010 GULF OIL SPILL
Katherine Rogers
Approved:
________________ _________________
Don W. Stacks, Ph.D. Shannon Campbell, Ph.D.
Professor of Public Relations Assistant Professor of Public
Relations
________________ ________________
Joseph Treaster, M.S. Terri A. Scandura, Ph. D.
Professor, John S. and James L. Knight Chair Dean of Graduate School
in Cross-Cultural Communication
ROGERS, KATHERINE (M.A., Communications)
British Petroleum’s Use of Image Restoration (May 2012)
Strategy on Social Media and Response After the
2010 Gulf Oil Spill
Abstract of a thesis at the University of Miami.
Thesis supervised by Professor Don Stacks.
No. of pages in text. (53)
As the population and use of social media platforms increase, so does the role
they play in crisis communication. This study focuses on how BP used social media,
specifically Twitter, as a platform for their public relations response to the devastating
2010 Gulf Oil Spill. Additionally, this study focuses on identifying how social media
users used Twitter as an avenue to reach out to BP on and voice concerns and attitudes
after the spill. A content analysis of Tweets written in response to the spill was conducted
and through that analysis it was determined that BP used Twitter as a way to disperse
messages about how they were making things better or compensating people; while social
media users used Twitter to attempt to engage with BP and voice negative opinions about
many areas that were damaged because of the spill. By identifying how BP used Twitter
and restoration strategies in the wake of the spill, it is possible to compare their use to
best practices that have developed through academic research and existing case studies
where social media and restoration strategy was used effectively. This can identify why
BP’s response was not well received. By identifying how social media users used
Twitter, it is possible to use this case study as an example of the importance of engaging
with users and identifying areas of concern. This study has implications for public
relations practitioners and all professionals who are charged with creating crisis
communication plans. Limitations of the study and recommendations for future research
were included.
iii
!"#$% '( )'*!%*!+
PAuE
LIST 0F TABLES .................................. iv
CBAPTER
1 INTR0B0CTI0N ............................ 1
2 LITERAT0RE REvIEW ........................ 4
0il Spill Cause anu Effects ......................... S
Legal Issues, Lack of Piepaieuness anu Impact on Clean-up .... 6
Similaiities between Exxon valuez anu BP Spill .......... 7
BP's Safety Recoiu anu 0se of Social Neuia .............. 8
Impact of 0il Spill on the Bealth of uulf Coast Citizens ........ 1u
Two Theoietical Ciisis Nouels ..................... 1u
Attiibution Theoiy anu Situation Ciisis Communication Theoiy . 11
Ciisis Communication Response Nouels .............. 1S
Coipoiate Communicative Response Nouel ............ 1S
A Changing Neuia Enviionment .................... 17
Social Neuia Revolution anu Ciisis Communication ........ 17
The Enviionmental Revolution ................... 2u
Reseaich Questions .......................... 21
S NETB0B0L0uY ............................ 22
0nits of Analysis ........................... 22
Sampling ................................ 22
Couing .................................. 2S
Couing Reliability ............................ 24
Statistical Analysis .......................... 24
4 RES0LTS ................................ 2S
Besciiptive Statistics .......................... 2S
Test of Reseaich Questions ...................... 27
S BISC0SSI0N ............................... S4
Summaiy ............................... S8
Implications ............................... S9
Limitations anu Biiection foi Fuithei Reseaich .......... 41
REFERENCES ................................... 4S
Appenuix A .................................. 46
Appenuix B .................................. S2
Appenuix C .................................. SS
iv
$,+! '( !"#$%+
Table 2-1: Attiibution Theoiy ........................... 12
Table 2-2: Benoit's (1997) Image Restoiation Stiategies ............. 14
Table 2-S: Biaufoiu anu uaiiet's (199S) Response Nouel ............. 16
Table 4-1: uenuei of Tweet Senuei ........................ 26
Table 4-2: Peicentage of Tweets Containing Restoiation Stiategy .......... 26
Table 4-S: Recoiueu Tone of Tweets Biiecteu Towaius BP ........... 28
Table 4-4: Recoiueu Focuses of Nessages ..................... 29
Table 4-S: Tone - Focus Cioss Tabulation ..................... Su
Table 4-6: uenuei - Tone Cioss Tabulation ................... Su
Table 4-7: Focus - Tone Cioss Tabulation ...................... Su
Table 4-8: Image Restoiation Stiategy Fiequencies ............... SS
1
Chaptei 1
Intiouuction
The oil spill in the uulf of Nexico that began on Apiil 2u, 2u1u, was the iesult
of an explosion anu fiie ioughly 4u miles off the coast of Louisiana. Biitish
Petioleum (BP) owneu anu opeiateu the Beepwatei Boiizon uiilling unit woiking
on an off shoie exploiatoiy well when malfunctions causeu the explosion. The
explosion claimeu the lives of 11 men woiking on the exploiatoiy well. As the fiie
buineu, oil pouieu fiom the well into the uulf of Nexico at an estimateu iate of
SS,uuu baiiels of oil pei uay. This continueu until }uly 1S, 2u1u, when the well was
finally cappeu. In total, between 18S anu 2uS million gallons of ciuue oil pouieu into
the uulf. In auuition to the tiagic loss of life, the spill pioveu to be uevastating to
maiine life in the uulf, killing anu enuangeiing countless animals anu entiie species.
uulf Coast inuustiies such as fishing, touiism, anu ieal estate iecoiueu uamages in
the billions of uollais. The spill took a physical anu psychological toll on those who
weie impacteu by its afteimath, both uiiectly anu inuiiectly.
Total uamages to BP, the enviionment, anu the economy in the uulf Coast aie
cuiiently estimateu to be $S6.9 billion. It was a uisastei on both human anu
enviionmental fionts anu Biitish Petioleum was left to figuie out how to combat all
aspects of the situation. BP nevei took uiiect iesponsibility foi the spill, iathei in
piess ieleases anu othei communications issueu following the explosion anu spill,
they focuseu on shifting blame anu assuiing the public they woulu honoi all
legitimate claims anu gave uetaileu infoimation about the steps that they weie
taking to stop the flow of oil anu iectify the situation. BP was faceu with a majoi
2
public ielations ciisis in the wake of this uisastei. They encounteieu many
challenges anu setbacks, incluuing insensitive iemaiks maue by CE0 Tony Baywaiu
anu complaints fiom the public that BP was not uoing enough iemeuy the situation.
The BP Beepwatei Boiizon 0il Spill began at the height of the enviionmental
movement. People weie incieasingly vigilant of enviionmental issues anu many
weie beginning to make a conscious effoit to become moie enviionmentally
iesponsible oi "gieen." Coipoiate social iesponsibility was becoming of gieat
impoitance anu consumeis consiueieu moie than just the piouuct anu piice when
making a puichase.
Piioi to the spill, oil itself was a contioveisial issue. uas piices weie soaiing
anu many Ameiicans weie taking a seiious financial hit because of the inciease.
With that being saiu, the issue of off shoie uiilling was becoming incieasingly
contioveisial even befoie the spill uue to conflicting conceins about the
enviionment anu iising cost of oil anu the iesulting incieaseu piice of gas. 0ff shoie
uiilling anu enviionmental policies playeu a pivotal iole in the 2uu8 piesiuential
election. Aftei the spill, Piesiuent Baiak 0bama hau to put in place ceitain
emeigency measuies anu policies, incluuing a moiatoiium on off shoie uiilling that
lasteu foi seveial months befoie being lifteu.
Fuithei, the spill occuiieu when social meuia was becoming incieuibility
impoitant anu influential in the meuia anu society. Euosomwan, Piakasan, Kouame,
Watson, anu Seymoui (2u11) state that "social meuia helps conveisations to ieach a
wiuei auuience leveiaging the 'long tail' concept, which means conveisations that
can be conveyeu to uiffeient foiums" (p. 8S). Social meuia platfoims staiteu
S
incieasing in ielevance in 2uuu. By }uly 2u1u, Facebook hau moie than Suu million
active useis, YouTube viueos weie ieceiving millions of views uaily, anu Twittei was
becoming incieasingly populai as a souice foi news anu communication
(Euosomwan et. al, 2u1u, pp. 81 - 84). News was tiaveling fastei thanks to social
meuia, anu anyone's opinion coulu be wiitten anu voiceu. Nany coipoiations anu
companies weie opeiating social meuia platfoims anu hau to have a plan foi theii
social meuia use in times of ciisis.
The BP oil spill anu its afteimath pioviue valuable lessons in ciisis
communication. It is a case that will seive as an example of what to uo anu what
shoulu be avoiueu when piepaiing a ciisis communication plan oi combating a
ciisis. The long teim effects of the spill aie not yet known, but in the shoit teim
theie aie lessons to be leaineu about not only what coulu have been uone to pievent
the spill but about how to combat an enviionmental ciisis on the public ielations
fiont. The puipose of this stuuy is to iuentify, specifically, the ciisis iesponse
stiategies employeu by Biitish Petioleum on Twittei anu why they may not have
been optimally effective in using social meuia to contiol the public ielations ciisis
that iesulteu fiom the explosion anu oil spill. Auuitionally, it will examine in uepth
how people iesponueu anu ieacteu on social meuia outlets.
4
Chaptei 2
Liteiatuie Review
In oiuei to examine the iole of ciisis communication stiategy anu the use of
social meuia as a platfoim to expiess opinion anu attituue in the BP 0il Spill
iesponse anu afteimath, theie aie seveial aieas of influence to be ievieweu.
0nueistanuing exactly how the oil spill was causeu anu the extent of the uamages
that iesulteu, incluuing uamages to the enviionment, economy, anu citizens' health
anu well-being pioviues eviuence as to why people weie so uismayeu by the spill
anu possessing the knowleuge that it coulu have been pieventeu with some basic
safety piecautions.
Legal issues weie plentiful aftei the spill. Questions of piepaieuness, guilt,
liability, anu iights of those effecteu abounueu aftei the uisastei. BP coulu have
lookeu to pievious enviionmental uisasteis anu how they weie hanuleu foi
guiuance as they moveu foiwaiu. The 1989 Exxon valuez oil spill in paiticulai
pioviueu a pieceuent as to how oil spills can be hanuleu in the 0niteu States. Ciisis
communication anu image iestoiation mouels (e.g., Benoit, 1997) aie in existence to
help oiganizations anu inuiviuuals piepaie foi anu iecovei fiom a ciisis. In auuition,
touay the social meuia must be auuiesseu in ciisis communication. Social meuia has
cieateu a sense of tianspaiency anu accessibility between citizens anu
oiganizations. A plan must auuiess the way it will be employeu in ciisis
management situations.
Finally, cuiient social tienus neeu to be taken into consiueiation uuiing
ciisis communication planning. The histoiy anu state of the "gieen" movement in
S
the 0niteu States anu the ability to know how it will impact ciisis management is
impoitant in the BP oil spill case because of the enviionmental uestiuction it causeu.
0il uiilling is a contioveisial enviionmental issue anu the spotlight on
enviionmental issues at the time of the spill maue the fallout even woise foi BP.
!"# %&"## '()*+ (,- .//+01*
An explosion on the Beepwatei Boiizon uiilling unit being opeiateu by
Biitish Petioleum ioughly 4u miles off the coast of Louisiana tiiggeieu the oil spill
that began on Apiil 2u, 2u1u, in the uulf of Nexico. The explosion was just the
beginning of a nightmaie foi BP, iesiuents of the uulf Coast, anu all those impacteu
by the spill. Issues of iesponsibility, accountability, liability, anu cause abounueu in
the wake of the spill as citizens of the uulf Coast watcheu theii enviionments anu
lives be giavely impacteu by the oil spilling into the uulf. Smith, Smith, anu Ashcioft
(2u1u) explain that the uamages fiom the spill aie
attiibuteu to thiee majoi factois: (1) human eiioi anu equipment failuie at
BP's Beepwatei Boiizon offshoie uiilling unit; (2) failuie of the 0.S.
goveinment to assign, anu in some cases to peimit, iesouices to assist with
the containment of the oil spill; anu (S) misinfoimation uisseminateu by the
news meuia iegaiuing the amount anu location of oil pollution in the watei
anu on the beaches of the uulf of Nexico. (p.1)
The iesults of the spill weie not easy to account foi, howevei; estimateu total
uamages to BP, the enviionment, anu the economy aie at least $S6.9 billion, though
this is subject to change. Some shoit-teim effects of the spill associateu with the
iesilience of the enviionment anu the oil spill aie known now, but the long-teim
effects will not be iealizeu foi yeais to come. The oil spill leu to laws anu iegulations
a moiatoiium on uiilling (which was latei iuleu aibitiaiy anu oveituineu) anu
iestiictions to aieas foi fishing (Smith et. al., 2u1u, p. S)!
6
2+3(# 4**)+*5 2(06 7/ 89+&(9+-,+** (,- 4:&(01 7, '#+(,
&
Along with the enviionmental anu humanitaiian uisastei that BP was
iesponsible foi taking caie of, they now hau to ueal with a massive public ielations
ciisis. Eviuence points to the fact that the ciisis coulu have been avoiueu all togethei
if piopei safety piecautions weie taken. Bowevei, the explosion happeneu anu oil
pouieu into the uulf at an alaiming iate. While theii piepaieuness to ueal with the
spill enviionmentally was being questioneu, theie weie also issues with whethei oi
not they hau a piopei ciisis communication plan in place. uiiggs (2u11), foi
example, iaises issues of piepaieuness. 0ne of the majoi issues that came into play
legally uuiing the spill was the pace of iesponse uue to the fact that BP woulu not
allow any piivate help without going thiough the lengthy peimission piocess. This
was a battle fought by local people who weie familiai with the aiea anu equippeu to
help in some ways, but weie piohibiteu fiom uoing so. This was ieflecteu in the
meuia as a souice of gieat fiustiation. At the ioot of this pioblem was the legal issue
that BP was financially iesponsible foi the cleanup anu theiefoie hau to appiove
eveiything. If they alloweu foi outsiue help befoie appioval they iiskeu the piivate
paity wanting to collect foi the woik they hau uone (uiiggs, 2u11, p. 7).
BP was not piepaieu foi the PR ciisis they woulu be uealing with the in the
afteimath of the spill. The Society foi Risk Analysis (SRA) <",";*=:&7*"): 7, 1>+
?++&@(1+9 A79"B7, !"# %&"## (2u1u) biought to question BP's piepaieuness on
seveial levels. In paiticulai, past SRA piesiuent Caion Chess questioneu BP's level
of iisk communication piepaieuness. Chess felt that BP lackeu any amount of
piepaieuness anu faileu miseiably fiom a iisk communications point of view.
7
Fuithei, Chess noteu similaiities between BP's iesponse anu Exxon's 1989 iesponse
to its oil spill in the Stiaits of valuez, Alaska. She citeu an aiticle fiom C9++,@"9+ that
noteu "uepiessing similaiities between Exxon's anu BP's iisk communication
smash-ups: shifting blame fiom the coipoiation, emphasizing science iathei than
people, anu initially uownplaying the potential impact of the uisastei" (Chess, 2u1u,
p. 216).
%":"#(9"1"+* D+1@++, .EE7, F(#-+B %&"## (,- G8 %&"##
BP was not without an example on how to ueal with spills. They hau plenty of
lessons that they coulu have taken away fiom the 1989 Exxon valuez oil spill. It is
impossible to uiscuss the iesponse to the BP oil spill without looking at othei majoi
inciuents that pieceueu it. The Exxon valuez oil spill seives as benchmaik of soits to
compaie the BP iesponse to, although the BP spill was laigei than the Exxon spill.
Ritchie et al. (2u11) note that one of the lessons leaineu fiom the Exxon spill is that
the BP spill iecoveiy is fai fiom ovei:
We know that social anu psychological stiess is heighteneu by the
unceitainty that comes fiom toxic contamination. The complexities of local
anu iegional economies anu the upheaval cieateu by a tempoiaiy economic
boom cieateu by cleanup opeiations also contiibute to the uisastei's
lingeiing effects. Finally, as injuieu paities seek compensation foi uamages
iesulting fiom this enviionmental uisastei, many will be tiaumatizeu by
impeisonal buieauciatic stiuctuies anu piotiacteu litigation. (p. S1)
Anothei similaiity is that both spills effecteu ienewable iesouice communities;
ienewable iesouice communities "uepenu on ienewable natuial iesouices foi theii
social, cultuial anu economic existence" (p. S1). Ritchie also uiaws similaiities
between those in the aiea of the spill who gaineu financially as a iesult of the spill
because of the cleanup effoits. Fuithei, uisasteis often uiive people to move fiom
8
theii communities. This iesults in economic loss anu human capital loss (Ritchie,
2u11, p. S4).
G8H* %(/+1= I+079- (,- J*+ 7/ %70"(# <+-"(
Buiing times of ciisis, a company's safety iecoiu anu image come into
question. Companies who have bettei ieputations going into a ciisis often come out
bettei than those who go into the ciisis unfavoiably. BP is no exception. The oil spill
has biought ciiticism fiom a wiue iange of people, fiom activist gioups to
goveinment entities. While BP has taken iesponsibility foi taking caie of the spill,
they have shifteu blame on subcontiactois Tiansocean anu Ballibuiton. A stuuy by
Nuialiuhaian, Billistone, anu Shin (2u11) choose to examine social meuia anu how
BP useu it uuiing the oil spill iesponse because it allows them to unueistanu "the
uimensions of auuience ieaction" which they claim to be "ciucial in a cycle of ciisis"
(p. 227). The social meuia seiveu as a platfoim foi the public to voice theii
conceins, issues, anu angei uuiing the spill. Nany times, the messages weie in
iegaius to BP's safety measuies in the past leauing up the spill anu how the spill
coulu have been pieventeu hau appiopiiate safety measuies been taken.
BP's safety iecoiu came into question initially in the eaily 2uuus. They came
unuei fiie even moie so aftei a 2uuS explosion at its Texas City, Texas, iefineiy that
claimeu the lives of 1S woikeis anu injuieu neaily 2uu otheis. Aftei the explosion,
BP assuieu the public that they hau leaineu fiom theii mistakes anu that safety was
a top piioiity. All of this was just five yeais piioi to the BP oil spill in the uulf of
Nexico Aftei the spill, much of BP's iesponses weie uefensive. In theii initial piess
ieleases, BP focuseu on how they weie the assisting paity iathei than the
9
iesponsible paity. The fiist piess ielease issueu on Apiil 21, 2u1u was a
confiimation anu copy of Tiansocean's ielease announcing the explosion (Biitish
Petioleum, p.1). BP's seconu piess ielease, also issueu on Apiil 21, 2u1u stateu that
BP touay offeieu its full suppoit to uiilling contiactoi Tiansocean Ltu.
Anu its employees aftei fiie causeu Tiansocean's submeisible uiilling
iig Beepwatei Boiizon to be evacuateu oveinight, saying it stoou
ieauy to assist in any way in iesponuing to the inciuent. (Biitish
Petioleum, paia. 1)
It was not until Apiil 24, 2u1u that BP even acknowleugeu publicly the loss of life
that occuiieu as a iesult of the spill (Biitish Petioleum, p. 1). BP iemaineu steaufast
that they weie in a suppoiting iole, anu not iesponsible foi the uisastei. In Bob
Buuley's Nay 6, 2u1u speech at the Chief Executives' Club of Boston (which was the
fiist speech aftei the spill), the event was blameu on a faulty blowout pieventei
(Buuley, 2u1u). As moie infoimation was ieleaseu, Ballibuiton's faulty cement job
woulu come into play as anothei guilty paity BP sought to blame. Eventually, it
came into question whethei oi not BP was awaie of the fact that the cement job on
the Beepwatei Boiizon well was faulty; with eviuence pointing to the fact that they
weie awaie. This was highly ciiticizeu by many people anu gioups. BP spent $Su
million uefenuing its iecoiu to the public. Nany felt that this was $Su million that
shoulu have been spent to clean up the spill anu aiu those who hau been affecteu.
Nuialiuhaian et al. (2u11) quote Chiis uiluez of the global public ielations fiim Bill
anu Knowlton as saying, "until the leak is stoppeu, no amount of auveitising anu PR
will help" (p. 227).
1u
4:&(01 7/ !"# %&"## 7, 1>+ A+(#1> 7/ C)#/ '7(*1 '"1"B+,*
Apiil 2u, 2u1u, will be ueeply etcheu into the minus of citizens of the uulf
Coast. The spill's impact on the enviionment anu economy was alieauy pioving to
be uisastious. Linua NcCauley (2u1u) states that citizens of the uulf Coast will be
affecteu by the oil spill with shoit anu long teim psychological uamage. In auuition
to inciuences of suiciue as a iesult of stiess causeu by the spill, iesponueis anu
iesiuents of the uulf aie at iisk foi ueveloping anxiety uisoiueis anu post-tiaumatic
stiess uisoiuei. Similai issues weie seen in the wake of the Exxon valuez oil spill.
NcCauley also cites eviuence that long-teim iesponueis have shown signs of
iiiepaiable psychological uamage (pp. S4 - S6). In the .,+93= K++6#= L+@* (2u11)
the psychological effects aie highlighteu again in moie uetail. This aiticle cites
stuuies anu statistics uemonstiating that a similai level of stiess was piesent in
inuiviuuals in uulf Coast states that weie uiiectly influenceu by the oil spill anu
those in states, such as Floiiua, that weie not as uiiectly impacteu. Accoiuing to
.,+93= K++6#= L+@* (2u11), a 0niveisity of Naiylanu anu 0niveisity of Floiiua
stuuy showeu that the oil spill was a huge public mental health ciisis anu suggests
that some iecoveiy effoits neeu to auuiess these issues (p. 1).
M@7 M>+79+1"0(# '9"*"* <7-+#*
Two mouels of ciisis iesponse may help to bettei unueistanu both BP's
iesponse stiategy anu the public's iesponse to that stiategy. Coombs' (2uu4)
Situational Ciisis Communication Theoiy examines the impact of piioi events anu
attiibutions baseu on those events on ciisis ieaction. Benoit's (1997) image
11
iestoiation appioach looks at the actual messages communicateu by an
oiganization aftei the ciisis has occuiieu.
N119"D)1"7, M>+79= (,- %"1)(1"7, '9"*"* '7:
,"0(1"7, M>+79=
Situational Ciisis Communication Theoiy (SCCT) posits that an inuiviuual's
knowleuge of past ciisis events an oiganization is chaigeu with being iesponsible
foi is necessaiy to take into consiueiation when auuiessing futuie ciisis
communication planning (Coombs, 2uu4, p. 26S). Accoiuing to Coombs', as
"ieputational thieat incieases, the ciisis managei shoulu use iesponse stiategies
that uemonstiate the acceptance of iesponsibility foi the ciisis anu that auuiess
victim conceins" (p. 266). BP, uue to the natuie of theii inuustiy anu the Texas City
iefineiy uisastei, was in a position of high ieputational thieat as they went into
ciisis iesponse situation uuiing the 2u1u oil spill. The high ieputational thieat anu
past iecoiu of safety issues leauing to seiious uisasteis means that BP woulu likely
be helu iesponsible in some uegiee foi the situation, no mattei how much they
coulu have influenceu oi pieventeu the situation.
Attiibution Theoiy is connecteu to SCCT because it holus that "people will
make juugments about the causes of events, especially unexpecteu events with
negative outcomes" (Weinei, 198S, p. 28u) anu that theie aie ceitain attiibutions, oi
peiceptions, that will influence how the event is juugeu. In paiticulai theie aie thiee
uimensions that aie useu to foim attiibutions, as iepiesenteu in Table 2-1.
12
Table 2-1
Attiibution Theoiy
#$$%&'($&)* +,-&*&$&)*
Stability If the cause happens fiequently (stable)
oi infiequently (unstable)
Exteinal Contiol Whethei the cause of the event was
contiollable oi uncontiollable by
anothei peison
Peisonal Contiol ¡ locus The extent to which a peison can contiol
his oi hei own fate in the situation
Each attiibution impacteu how inuiviuuals juugeu BP befoie, uuiing, anu aftei the
oil spill. Stability was uebateu because of the uisastei that claimeu seveial lives just
S yeais pievious to the spill. Exteinal contiol was a majoi factoi uuiing the post-
uisastei sciamble to finu facts about what happeneu anu whethei oi not it coulu
have been pieventeu. The uistinctiveness of the situation was calleu to question as
BP's tiack iecoiu was examineu. Accoiuing to Coombs (2uu7), a ciisis is not
uistinctive when "an oiganization shows little consiueiation foi stakeholueis acioss
a numbei of uomains, not just in ciisis" (p.1S7). As moie infoimation became cleai
about the uetails of what happeneu anu BP was woiking to iesponu to the public
ielations ciisis, manageis shoulu have lookeu to Situational Ciisis Communication
Theoiy as a guiue to finuing the best possible way to iesponu, given theii ieputation
anu the situation.
'9"*"* '7:,"0(1"7, I+*&7,*+ <7-+#*
Theie aie two mouels that attempt to explain ciisis communication
stiategies anu iesponses. Benoit (1997) takes a moie ihetoiical look at the ciisis
1S
communication messages by focusing on the stiategic iationale behinu the message,
while Biaufoiu anu uaiiet (199S) look at a two-factoi appioach that focuses on a
iesponse oi no iesponse uichotomy.
4
3+ 9+*179(1"7, 1>+79=
BP was left with the uaunting task of iestoiing its image to tiy anu get back
in the public favoi. This is a veiy uifficult public ielations task, especially when
uealing with a uisastei as laige anu uevastating as the oil spill. Benoit's (1997)
"Image Restoiation Theoiy" pioviues the theoietical ieasoning behinu why ceitain
stiategies aie best employeu in paiticulai situations. The theoiy is baseu on basic
concepts that piopose that the accuseu paity is iesponsible anu that the accusation
is consiueieu offensive (Benoit, 1997, p. 177). Table 2-2 piesents the 14 stiategies
that have emeigeu fiom Benoit's theoiy iange fiom simple uenial to moitification
(Bailow, 2u11, p! 81).
0sing Benoit's theoiy as a basis foi his stuuy, Bailow says that "BP was
focuseu on the twin tasks of uesciibing coiiective action anu uesciibing attempts to
compensate the victims of the inciuent" (p. 82). Fuithei, he states that BP nevei
boie any of the iesponsibility foi the actual explosion anu iesulting oil spill, iathei
they simply accepteu iesponsibility foi uealing with its afteimath. This is a point
that causeu significant tension (p. 82).
14
Table 2-2
Benoit's (1997) Image Restoiation Stiategies
.$%/$,01 23/456,
Simple uenial We uiu not uo it
Shifting blame Someone else uiu it
Piovocation We uiu it, but weie piovokeu
Befeasibility Lack of infoimation oi contiol
Acciuents The inciuent was an acciuent
uoou intention The eiioi was the iesult of goou intentions
Bolsteiing 0ui goou chaiacteiistics outweigh any eiioi
Ninimization The pioblem is not that bau
Biffeientiation The inciuent is uiffeient fiom some othei act
Tianscenuence
The act shoulu be unueistoou in a uiffeient
context
Attack accusei The peison blaming us is the one at fault
Compensation The victim will be compensateu
Coiiective Action We will fix the pioblem
Noitification
We aumit iesponsibility oi ask foi
foigiveness
1S
'79&79(1+ 07:,"0(1"O+ 9+*&7,*+ :7-+#
Ciisis communication is a uelicate task to hanule. Theie must be a ciisis
communication plan in place piioi to a ciisis occuiiing in oiuei to piopeily auuiess
anu combat the situation. Biaufoiu anu uaiiet's (199S) coipoiate communicative
iesponse mouel "focuses on the potential effectiveness of coipoiate communicative
iesponses to accusations of unethical oiganizational behavioi" (p. 876). Theii
mouel is compiiseu of two paths: the no iesponse path anu the coipoiate iesponse
path. The no iesponse path finus that uuiing times of ciisis, oiganizations in
question may offei no iesponse foi the allegations being biought. Biaufoiu anu
uaiiet aigue that the no iesponse path geneially leaus to obseiveis assuming guilt);
alteinatively, the coipoiate iesponse path holus that, by offeiing some iesponse, an
oiganization can influence the obseivei's peiception of the situation (p. 878). 0nuei
the coipoiate iesponse path falls foui iesponses that may be offeieu in times of
ciisis. Table 2-S piesents these iesponses anu the optimal situations foi theii use.
16
Table 2-S
Biaufoiu anu uaiiet's (199S) Response Nouel
7,85)*8, +,-&*&$&)* .&$(/$&)*
Benials Benials aie statements that
ueny the occuiience oi
existence of the
questionable event, oi ueny
that the accuseu
oiganization is the cause of
the event.
When an actoi can pioviue
eviuence that he¡she uiu not
commit an allegeuly unethical
action (commission situation),
this actoi shoulu use a uenial
communicative iesponse.
Excuses Excuses aie statements that
aigue that the accuseu
oiganization shoulu not be
helu iesponsible foi the
occuiience anu¡oi impact
of the questionable event
because ceitain factois
limiteu the oiganization's
contiol ovei the occuiience
anu¡oi impact of this event.
When an accuseu actoi can
pioviue eviuence that he¡she
uiu not have contiol ovei the
occuiience anu¡oi impact of
an allegeuly unethical action
(contiol situation), this actoi
shoulu use an excuse
communicative iesponse
}ustifications }ustifications aie statements
that aigue that, even though
the accuseu oiganizations is
iesponsible foi the
questionable event, the
stanuaius being useu by the
accuseis to evaluate the
impact of this questionable
event aie inappiopiiate.
When an accuseu actoi can
pioviue eviuence that
inappiopiiate stanuaius aie
being useu to evaluate this
allegeuly unethical action
(stanuaius situation), this
action shoulu use a
justification.
Concessions Concessions aie statements
that agiee that the
questionable event uiu
occui, that the accuseu
oiganization causeu this
event, that they evaluative
stanuaius being useu by the
accuseis aie appiopiiate.
When an accuseu actoi
concluues that the allegations
iaiseu aie valiu (agieement
situation), the actoi shoulu
use a concession.
Bue to the scope anu natuie of the BP uisastei, issuing a iesponse of some natuie
was necessaiy. Bowevei, in the beginning of theii communicative iesponse, BP uiu
17
what they coulu to tiy to shift the blame to othei playeis in the uisastei. This mouel
uemonstiates that the iesponse BP chose uiu not coiiesponu with the situation
cieateu by theii unethical behavioi. This likely playeu a iole in the level of public
acceptance of the communicative anu public ielations measuies BP took in the wake
of the uisastei.
N '>(,3",3 <+-"( .,O"97,:+,1
Two factois meuiate any analysis of BP's ciisis iesponse stiategy anu
communications. The fiist ueals with a changing communication meuium÷the iise
of a populist-appioacheu social meuia. The seconu ueals with a ieneweu awaieness
of the fiagility of the woilu's enviionment. Both help to bettei unueistanu BP's
stiategy anu its failuies at communicating.
%70"(# :+-"( 9+O7#)1"7, (,- 09"*"* 07:,"0(1"7,
The BP oil spill happeneu uuiing a time when two veiy impoitant
ievolutions weie happening. Fiist, social meuia was taking ovei anu changing the
way communications, public ielations, anu oiganizations functioneu. Wateis,
Tinuall, anu Noiton (2u1u), foi instance, examineu how public ielations has
iesponueu to anu evolveu with the incieaseu use of social meuia. They founu that
moie iepoiteis aie now open to getting news anu pitches thiough alteinative
souices. The stuuy goes so fai as to cite }ulia Boou, then euitoi at 8I K++6, as saying
that the "meuia pitch is ueau" (p. 242).
Reaching the meuia anu public is changing because of social meuia, foicing
iepoiteis, publicists, anu all communicatois to auapt. The meuia is no longei solely
18
getting stoiies thiough tiauitional pitches; iathei they aie looking at alteinative
methous, incluuing social meuia, to get stoiies. This is impoitant when stuuying the
BP oil spill because it was one of the fiist expeiiences with ciisis anu social meuia.
Buiing the Exxon valuez oil spill, public ielations piactitioneis uiu not have to
woiiy about Twittei anu Facebook anu Blogs. The BP oil spill bioke the molu anu
public ielations piofessionals hau to seiiously consiuei the iepeicussions of social
meuia anu how they woulu utilize it to combat the situation. It was also the fiist oil
spill wheieby the public hau access to social meuia. Repoiteis weie able to ieach
out the public on social meuia to get feeuback anu infoimation.
Social meuia, as it is available at peoples' fingeitips, has become influential in
how an inuiviuual foims opinions. Twittei, in paiticulai, has cieateu a sense of
community wheie inuiviuuals use 14u chaiacteis to post links, expiess opinions,
anu shaie infoimation with those who "follow" them anu aie pait of theii
community. The community peimits the shaiing of opinions anu values between
useis (Auei, 2u11, p. 711). Fuitheiing this, stuuies conuucteu just aftei the 2uu8
Piesiuential election in the 0niteu States showeu that among social meuia useis
ageu 18 - 29, Su% of the population polleu by Pew saiu they useu social meuia to
uiscovei, leain, anu influence theii voting uecision anu opinions on issues. Foi useis
age Su anu ovei, the peicentage was S2% (Auei, 2u11, pp. 711-712).
The piopensity of people to use social meuia as a place to gathei news uuiing
ciisis can be uemonstiateu by compaiing the iesponse of a pie-social meuia ciisis to
a cuiient ciisis. Buiing the teiioiist attacks of Septembei 11, 2uu1, officials hau to
spenu millions of uollais anu time to synchionize communication systems. In
19
compaiison, uuiing the Baiti eaithquake of 2u1u, Twittei was useu as a meaningful
channel of emeigency communication (Auei, 2u11 ,p.72S).
Social meuia has also incieaseu as souices foi news because of the gieatei
amount of fieeuom it allows foi. Wheie ceitain types of images oi stoiies aie
iestiicteu, social meuia has bioken the baiiiei anu alloweu foi all news anu stoiies
to be given aiitime (Cheong anu Lee, 2u1u, p. 47). This was uemonstiateu most
iecently uuiing the so-calleu "Twittei Revolution" anu othei social meuia leau
ievolts (Noiozov, 2u1u, pp. 24 - 2S).
Twittei has become a poweiful woiu of mouth tool to shaie sentiments anu
expeiiences, both positive anu negative, about bianus anu companies. }ansen,
Zhang, Sobel, anu Chowuuiy (2uu9) foi example founu that 19% of micioblogs
containeu a mention of a bianu oi company, anu of that peicentage, Su% weie
positive anu SS% weie negative (p. 2169). Twittei, as a micioblog, is accessible at
all times foi a usei to quickly post an opinion as it comes to top of minu. It enables
useis to immeuiately wiite how they feel oi what they think uuiing expeiiences.
Theii stuuy also founu that theie is a 0-shapeu ielationship between an inuiviuual's
expeiience anu theii likelihoou to expiess it (}ansen et al., 2uu9, p. 2178).
Twittei is aichival, which incieases its ielevance anu impoitance in
communication. Although theie aie limits to this on the actual Twittei website,
theie aie numeious ieliable aichive websites wheie you can access the past
(Fischei, 2u1u, p. 1S). 0nlike a face-to-face conveisation, wheie uetails can be lost
ovei time, Tweets can be founu anu tiaceu back anu founu in the context that they
weie wiitten. They have become pait of histoiy in how they have been useu in ciisis
2u
anu leau to ievolution. Social netwoiks, incluuing Twittei, have uemonstiateu theii
impoitance by allowing foi the continuation of piogiess anu thiough pioviuing an
"instantly accessible anu ongoing inventoiy of infoimation that can be useu to solve
pioblems anu even potentially to pioviue emotional suppoit" (NcSwite, 2uu9, p.
92).
M>+ .,O"97,:+,1(# I+O7#)1"7,
The seconu ievolution taking place uuiing the uulf oil ciisis was a shift of
focus towaius enviionmentalism. Caie foi the enviionment anu instiuction to uo so
can be tiaceu back to ancient times. It is in the G"D#+ anu P)9H(,. Theouoie Roosevelt
leu an enviionmental movement uuiing his piesiuency between 19u1 anu 19u9.
Although its histoiy uates back so fai, many people uiu not take into consiueiation
how theii actions effect the enviionment until moie iecently. Enviionmentalism has
become an impoitant pait of business anu especially maiketing uue to the fact that
costumeis aie now looking foi enviionmental anu social iesponsibility when
making puichases. Stuuies have founu that consumeis piefei "gieen" piouucts anu
woulu switch bianus anu spenu moie money in oiuei to puichase these piouucts
(Smith, 2u1u p. 2). Even piioi to the oil spill, the uamages causeu by oil anu gas
companies tuineu people off to these inuustiies anu theii images weie huiting.
Thus, the uegiee of ciisis facing BP was enhanceu uue to a ieneweu iesponse to
enviionmental uisasteis anu impacteu on not only what people communicateu
about the ciisis but also in new communication channels.
21
I+*+(90> P)+*1"7,*
Nost of the existing liteiatuie iegaiuing the 2u1u uulf 0il Spill examines the
causes, effects, anu outcomes of the uisastei. The complete outcome of the spill is
unknown anu may not be fully unueistoou foi sometime as inuustiy anu
enviionment both iequiie significant time to bounce back, anu fact finuing anu legal
battles iegaiuing the spill aie still ongoing. Existing social meuia liteiatuie laigely
focuses on pievalence of uiffeient platfoims anu how the powei of social meuia can
be hainesseu anu best useu to solve ceitain communication issues. uiven the extent
of existing ieseaich on social meuia anu liteiatuie iegaiuing ciisis communication,
it is possible to stuuy how ceitain stiategies weie employeu in BP's public ielations
iesponse following the spill. Baseu on the pieceuing uiscussion, the following
ieseaich questions aie poseu:
IPQR What was the uominant tone of Tweets uiiecteu towaius Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
IPSR What was the uominant focus of Tweets uiiecteu towaiu Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
IPTR What weie the uominant Image Restoiation Stiategy employeu by
Biitish Petioleum on Twittei in iesponse to the 2u1u uulf oil spill.
22
Chaptei S
Nethouology
In oiuei to stuuy both how Biitish Petioleum useu Image Restoiation
Stiategy (Benoit, 1997) in theii social meuia iesponse to the oil spill anu how the
public useu social meuia as a platfoim to iesponu anu ieach BP aftei the 2u1u uulf
oil spill, a content analysis was conuucteu. The analysis of Tweets geneiateu by
BP_Ameiica anu Tweets wiitten by the public to ÇBP_Ameiica weie all piouuceu
between Nay 2, 2u1u anu }anuaiy 1S, 2u12. The gieatest peicentages (89.S% anu
8S.S%) of content came fiom the fiist few months immeuiately following the oil
spill.
J,"1* 7/ N,(#=*"*
Seveial units of analysis weie analyzeu in oiuei to gathei all necessaiy
infoimation to answei the ieseaich questions. The units of analysis incluueu social
meuia type, message tone, genuei of senuei, uate of message, message focus, anu BP
image iestoiation stiategy (see Appenuix A foi Couebook).
%&#",3
All Tweets useu in the samples foi content analysis weie selecteu at ianuom.
In oiuei to iuentify the vaiious image iestoiation stiategies useu by BP, a sample of
1u9 Tweets was selecteu at ianuom fiom the initial sample of Suu Tweets gatheieu
fiom an aichive seivice. A seconu sample on 82 Tweets uiiecteu at BP was
ianuomly selecteu fiom the initial gatheieu sample of 17u Tweets uiiecteu at BP
fiom an aichive seivice. In all, 191 Tweets compiiseu the sample useu to analyze
BP's iestoiation stiategies anu answei the thiee stateu ieseaich questions.
2S
'7-",3
Tweets piouuceu by the public uiiecteu towaius ÇBP_Ameiica weie coueu
foi the following vaiiables: uate (mm¡uu¡yy), genuei of the senuei, tone of the
message, anu focus of the message. Tweets piouuceu by BP_Ameiica weie coueu foi
the following vaiiables: uate (mm¡uu¡yy), image iestoiation¡not image iestoiation,
anu Benoit's 14 image iestoiation stiategies (See Table 2-2). Each vaiiable anu its
meaning was agieeu upon by the coueis (See Appenuix A ):
• +/$,: The yeai, month, anu uate that each inuiviuual tweet was
publisheu. Each Tweet in this stuuy hau a uate that shows when it
was publisheu.
• 9,*:,%: When possible, the genuei of the inuiviuual was iecoiueu as
"male" oi "female". At times, the genuei was uncleai anu iecoiueu as
"othei".
•*,: When applicable, the Tweets weie analyzeu foi tone. Content
that was wiitten in suppoit of BP oi its effoits to clean the uulf,
iectify the situation, oi take iesponsibility foi what they hau uone
weie classifieu as "positive". Content that conuemneu BP, its
measuies in the uulf, theii iesponse to the oil spill in geneial weie
categoiizeu as "negative". Also incluueu as negative weie any Tweets
that ciiticizeu BP's iesponse as ineffective, insufficient, oi impiopei.
Content that challengeu BP's honesty anu integiity in the cleanup oi
afteimath in any weie categoiizeu as negative. Some content
containeu both positive anu negative iemaiks towaius BP anu theii
24
iesponse to the oil spill anu as such was categoiizeu as "neutial".
Content that expiesseu no stiong feeling, offeieu suggestions oi
claiifications was also categoiizeu as neutial.
• <)=(8 )- $>, ?,88/0,: When possible, the coueis iuentifieu the main
focus of the Tweet as being in iegaius to inuustiy uamage oi
piogiess, health conceins, enviionmental uamage oi piogiess,
financial anu economic impact of the spill, iecoveiy effoits, oi othei.
• @4/0, 7,8$)%/$&)* .$%/$,01: When applicable, the coueis
categoiizeu the Tweets into one of the 14 uiffeient stiategies
incluueu in Benoit's (1997) Image Restoiation Stiategy theoiy (See
Table 2-2).
'7-",3 9+#"(D"#"1=
The ieseaichei anu anothei public ielations giauuate stuuent, aftei agieeing
on guiuelines anu instiuctions, coueu a test set of 1u7 Tweets uiiecteu to
ÇBP_Ameiica anu geneiateu by ÇBP_Ameiica anu ieacheu an acceptable level (.84)
of ieliability ueteimineu by using Bolsti's (1969) ieliability foimula.
%1(1"*1"0(# N,(#=*+*
Aftei couing ieliability was establisheu, the sample sets weie coueu anu all
uata was enteieu into SPSS foi uesciiptive anu infeiential statistical analyses.
Specifically, uata weie uisplayeu by fiequency anu peicent anu testeu foi
uiffeiences via Chi-squaie tests.
2S
Chaptei 4
Results
The puipose of the specific ieseaich questions that weie put foiwaiu in this
stuuy was to specifically inuentify how iestoiation stiategy was useu on Twittei,
how inuiviuuals useu Twittei as a platfoim to voice theii opinions in the wake of the
spill, anu finally to ueteimine the piimaiy concein voiceu by of Twittei useis aftei
the spill. Answeiing these questions pioviues bettei insight into how social meuia
can be best employeu in ciisis anu why BP's ciisis communication plan uiu not woik
out that well.
The stuuy conuucteu a content analysis of 191 Tweets geneiateu by oi
uiiecteu at Biitish Petioleum aftei the Apiil 21, 2u1u explosion on the Beepwatei
Boiizon uiilling iig anu the iesulting oil spill. All uata collecteu fiom the sample set
of Tweets weie analyzeu using Statistical Package foi the Social Science (SPSS).
?+*09"&1"O+ %1(1"*1"0*
The uata collecteu to answei the ieseaich questions was gatheieu fiom the
Twittei aichive seivice Topsy. 17u Tweets ÇBP_Ameiica weie collecteu to gathei a
sample size of 82 Tweets foi analysis. Analyzeu content came fiom uates beginning
on Nay 2, 2u1u, anu enuing on Septembei 14, 2u11. Although the Tweets foi the
analyzeu sample weie collecteu at ianuom, the majoiity of content (8S.S%) was
geneiateu in 2u1u fiom the beginning of the oil spill until August 2u1u, just aftei the
well was cappeu. In the afoiementioneu sample the genuei of the senueis was
iepoiteu as SS.4% male, 17.1% female, anu 47.6% othei (see Table 4-1).
26
Table 4.1 uenuei of Tweet Senuei
Gender
Frequency Percent Valid Percent
Cumulative
Percent
Valid Other 39 47.6 47.6 47.6
Male 29 35.4 35.4 82.9
Female 14 17.1 17.1 100.0
Total 82 100.0 100.0
Bata collecteu to answei the ieseaich questions in iegaiu to Tweets
geneiateu by Biitish Petioleum weie also gatheieu fiom Twittei aichive seivice
Topsy. Thiee hunuieu twenty Tweets fiom BP_Ameiica weie collecteu anu a sample
of 1u9 Tweets was selecteu at ianuom foi analysis. Analyzeu content ianges in uate
fiom Nay 1u, 2u1u, to }anuaiy 1S, 2u12. Although the Tweets weie selecteu at
ianuom, in the same mannei as the fiist sample, the majoiity (89.S%) of the Tweets
in the sample aie fiom the time span fiom the beginning of the spill to shoitly aftei
the well was cappeu. Almost eighty peicent (79.8%) of the content in the sample
was iestoiation stiategy, while 2u.2% of the content was not iestoiation stiategy
(see Table 4-2). Tweets coueu as not iestoiation stiategy simply containeu some
basic infoimation oi uiu not peitain to the oil spill anu iecoveiy.
Table 4.2 Peicentage of Tweets Containing Restoiation Stiategy
Frequency Percent Valid Percent
Valid Restoration 87 79.8 79.8
Not Restoration 22 20.2 20.2
Total 109 100.0 100.0
27
M+*1 7/ I+*+(90> P)+*1"7,*
IP QR What was the uominant tone of Tweets uiiecteu towaius Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
In oiuei to ueteimine the uominant tone of the Tweets uiiecteu towaius BP,
each inuiviuual unit of analysis was coueu to ueteimine whethei the oveiall tone of
the message was positive, negative, oi neutial (X
2
=47.7S, uf=2, p=u). As founu in
Table 4-2, iesults showeu that of the 1u9 tweets analyzeu that 6S.4% (S2) weie
negative in tone. The uefinition of "negative" as useu in the analysis was any content
that conuemneu Biitish Petioleum, oi any content that ciiticizeu BP's iesponse as
ineffective, insufficient, oi impiopei. Foi example, negative Tweets incluueu one
fiom Nay 19, 2u1u when Çgeoffliving Tweeteu "It seems like ÇBP_Ameiica claim to
Sk baiiels a uay in the oil spill is the biggest PR lie of the yeai so fai. #BPheaiing".
Neutial-toneu Tweets accounteu foi SS.4% (29) of the analyzeu content.
"Neutial" was uefineu as content that containeu an equal amount of positive anu
negative, content that uiu not expiess stiong feeling, anu content that gave
suggestions, claiifications. Content coueu as neutial incluueu a Tweet fiom Nay S,
2u1u by Çzaibatsu that says "Let ÇBP_Ameiica know the #0ilSpill is impoitant to
you" anu anothei Tweet fiom }uly 1, 2u1u when PBS News Boui (ÇNewsBoui)
Tweeteu, " Watch ÇBP_Ameiica inteiview live on oui Çustieam". The final 1.2%
(1) of the content was positive in tone. In oiuei foi a Tweet to be "positive," it hau to
contain content that was in suppoit of BP anu theii effoits to clean the uulf oi in
28
piaise of the piogiess maue by BP in the uulf. The only iecoiueu positive Tweet was
a Tweet that stateu that BP hau biought the community togethei in the uulf foi a
laige baibeque event.
Table 4.S Recoiueu Tone of Tweets Biiecteu Towaius BP
Frequency Percent Valid Percent
Valid Negative 52 63.4 63.4
Neutral 29 35.4 35.4
Positive 1 1.2 1.2
Total 82 100.0 100.0
Chi-squaie value: 47.7S
IP SR What was the uominant focus of Tweets uiiecteu towaiu Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
To ueteimine the uominant focus of Tweets uiiecteu towaiu Biitish
Petioleum in the uata set, content was coueu anu categoiizeu in one of six possible
categoiies, incluuing: inuustiy, health, enviionment, financial anu economic impact,
iecoveiy, anu othei. The puipose of this ieseaich question is to ueteimine what the
uominant conceins uuiing the oil spill weie among inuiviuuals anu gioups who
useu Twittei to communicate conceins anu issues uuiing the oil spill anu clean up
(see Table 4-4).
"0thei" accounteu foi the laigest peicentage, S9% (S2) of content. Nany of
the Tweets that fell into this categoiy containeu messages that expiesseu fiustiation
as whole towaius BP, iathei than conceining a specific focus. The enviionment was
the most focuseu on specific categoiy, with Su% (2S) of the tweets containing
content that conceineu what BP was uoing to fix the enviionment oi othei content
29
that was in iegaius to enviionmental impact uue to the spill. Tweets iegaiuing the
iecoveiy piocess compiiseu 2u.7% (17) of the analyzeu content. 4.9% (4) of Tweets
weie in iegaiu to the financial anu economic impact that the spill hau on those that
it hau affecteu. Inuustiy anu health both weie the focus of 2.4% (2) of the content
that was analyzeu to answei the above stateu ieseaich question.
Table 4.4 Recoiueu Focuses of Nessages
Frequency Percent Valid Percent
Valid Industry 2 2.4 2.4
Health 2 2.4 2.4
Environment 25 30.5 30.5
Financial / Economic impact 4 4.9 4.9
Recovery 17 20.7 20.7
Other 32 39.0 39.0
Total 82 100.0 100.0
Chi-squaie value: 61.S6, uf=S, p=u
To expanu on the fiist two ieseaich questions, cioss tabulation was
calculateu to ueteimine any iiiegulai oi extieme vaiiances between tone anu focus,
genuei anu tone, anu genuei anu focus (see Tables 4-6, 4-6, anu 4-7).
Su
Table 4.S Tone - Focus Cioss Tabulation
Focus
Total Industry Health Environment
Fin. and econ.
impact Recovery Other
Tone Negative 1 1 20 3 10 17 52
Neutral 1 1 5 1 6 15 29
Positive 0 0 0 0 1 0 1
Total 2 2 25 4 17 32 82
Chi-squaie value: 8.928
Table 4.6 uenuei - Tone Cioss Tabulation
Tone
Total Negative Neutral Positive
Gender Other 23 15 1 39
Male 21 8 0 29
Female 8 6 0 14
Total 52 29 1 82
Chi-squaie value: 2.SuS
Table 4.7 uenuei - Focus Cioss Tabulation
Focus
Total Industry Health Environ.
Fin. and econ.
impact Recovery Other
Gender Other 1 1 13 2 10 12 39
Male 1 1 9 1 2 15 29
Female 0 0 3 1 5 5 14
Total 2 2 25 4 17 32 82
Chi-squaie value: 8.2S
The cioss tabulation tables uemonstiate that theie is vaiiance between tones
anu focus (X
2
=8.9S, uf=1u, p=.SSS). The aieas of focus that weie most pievalent in
the sample also show the highest concentiation of negativity. Tweets piouuceu by
S1
females appeai to be moie evenly split between negative anu neutial content than
Tweets piouuceu by males (X
2=
2.S1, uf=4, p=.64)
. The genuei - focus cioss
tabulation (X
2=
8.2S, uf=1u, p=.6u4) shows some unbalance in the uistiibution of
message focus between the genueis. Women show a highei concentiation of Tweets
focuseu on iecoveiy while men hau a highei focus on Tweets focuseu on the
enviionment. The highei peicentage of males in the sample, in auuition to the high
fiequency of Tweets that weie unable to be categoiizeu into one of the five aieas of
focus likely may have skeweu the iesults in the paiticulai aiea.
IP TR What weie the uominant Image Restoiation Stiategy employeu by
Biitish Petioleum on Twittei in iesponse to the 2u1u uulf oil spill.
0ne hunuieu nine Tweets geneiateu by Biitish Petioleum weie analyzeu to
ueteimine the uominant Image Restoiation Stiategy (Benoit, 1997) employeu by BP
in Twittei content aftei the spill anu uuiing the iecoveiy fiom the uisastei
(X
2
=169.26, uf=9, p=u). As seen in Table 4-7, almost 8u% (87) of the Tweets weie
iestoiation stiategies anu able to be categoiizeu in nine of Benoit's possible 14
stiategies. 2u% (22) Tweets weie not image iestoiation, meaning that they weie
just some foim of simple infoimation anu unable to be categoiizeu in one of the
possible 14 stiategies. Foi example, on }uly 7, 2u1u ÇBP_Ameiica Tweeteu an
announcement encouiaging followeis to "stay tuneu foi live tweeting of the latest
upuates." This Tweet is just a way of infoiming followeis that moie infoimation will
be ieleaseu in the neai futuie, it is not a iestoiation tactic. Anothei example of
S2
content that was coueu as not iestoiation is fiom }uly 14, 2u1u when ÇBP_Ameiica
Tweeteu about a biiefing being moveu up by an houi.
Foity-two peicent (46) of the Tweets weie coueu as coiiective action, which
was the most common stiategy employeu by BP. Coiiective action Tweets focuseu
on technical measuies that weie being taken by BP to stop the spill anu clean the oil.
0n }uly 1S, 2u1u ÇBP_Ameiica Tweeteu "At 12:Supm we staiteu to close the choke
- we uiu it half a tuin at a time to slowly close the well in." This Tweet was coueu as
coiiective action anu is fiom the uay that the well was finally sealeu. Anothei
example of coiiective action was Tweeteu on Nay 1u, 2u1u when ÇBP_Ameiica
saiu "We aie also in the piocess of gaining access at the bottom of the B0P. The next
step is the junk shot anu top kill to shut off the flow."
Bolsteiing was the seconu most common stiategy anu encompasseu almost
16% (17) of the coueu content. Bolsteiing was uefineu as when BP tiieu to
emphasize the positive things they weie uoing iathei than the negative events that
weie actually taking place. Foi example, BP often Tweeteu about seafoou safety anu
how veiy few tests on uulf sea foou weie coming back with positive iesults foi
contamination. This iecuiient theme uemonstiates bolsteiing because while all the
seafoou was supposeuly safe, the well was still not capeu anu oil was still flowing.
0f the iest, 9.2% (1u) weie classifieu as using compensation as iestoiation
stiategy; tianscenuence was employeu in 4.6% (S) of the analyzeu uata;
minimization as stiategy was useu in 2.8% (S) in the uata; shifting blame anu goou
SS
intention weie each obseiveu 1.8% (2) of the time; anu simple uenial anu attacking
the accusei weie each employeu in .9% (1) of the content .
Table 4.8 Image Restoiation Stiategy Fiequencies
Frequency Percent Valid Percent
Valid Not Restoration Strategy 22 20.2 20.2
Simple Denial 1 .9 .9
Shifting Blame 2 1.8 1.8
Good Intention 2 1.8 1.8
Bolstering 17 15.6 15.6
Minimization 3 2.8 2.8
Transcendence 5 4.6 4.6
Attack the Accuser 1 .9 .9
Compensation 10 9.2 9.2
Corrective Action 46 42.2 42.2
Total 109 100.0 100.0
Chi-squaie value: 169.26
S4
Chaptei S
Biscussion
Social meuia has become incieasingly impoitant in the past uecaue as moie
platfoims have become available foi useis to connect with othei useis anu geneiate
theii own content. It is now expecteu that an inuiviuual, oiganization, oi business be
piesent anu active to some uegiee on social meuia platfoims. Thiough cieating
connections between inuiviuuals anu oiganizations, social meuia has changeu the
business climate. An inuiviuual can now seek costumei seivice via social meuia
iathei than the olu fashioneu way of making a phone call. In that way anu some
otheis, social meuia has incieaseu the efficiency of oiganizations anu connecting
with the public. Bowevei, it has also alloweu foi inuiviuuals to have a platfoim to
spieau messages of uiscontent oi bau expeiiences to all of theii "followeis" anu
"fiienus" anu those Tweets, posts, anu viueos have uemonstiateu theii ability to
cause so much uamage by going viial oveinight. Anothei change social meuia is in
pait iesponsible foi is the call foi moie tianspaiency. Inuiviuuals now uemanu to
know how a company is opeiating anu how theii policies will affect them, anu social
meuia has cieateu a connection between inuiviuual anu coipoiations to facilitate
this exchange in infoimation.
In Apiil 2u1u, when oil began to poui into the uulf anu ieach the shoies in
numeious coastal communities, the effecteu inuiviuuals, inuustiy leaueis, anu
auvocates weie iightly infuiiateu with Biitish Petioleum. The tiuth about who is to
blame foi the spill anu the total uamage is still unknown anu evolving as moie
infoimation is ieleaseu anu stuuies aie conuucteu. Theie aie many lessons to be
SS
leaineu fiom the oil spill, anu one of those lessons is about the use of social meuia in
times of ciisis. Inuiviuuals can ieach an oiganization anu voice conceins,
fiustiations, anu suggestions in a public foium like nevei befoie. Similaily, Biitish
Petioleum was able to use social meuia as a platfoim to auuiess conceins anu
fiustiations while ieaching mass amounts of people with infoimation to help
contiol the ciisis, if uone piopeily.
This stuuy focuseu solely on the social meuia platfoim Twittei anu how
inuiviuuals useu it to connect with BP anu how BP, in tuin, useu it to ieach
inuiviuuals anu tiy to iecovei fiom the public ielations ciises that ensueu as a iesult
of the explosion anu spill. Twittei allows foi a iapiu way to engage because it's
accessible anu fiee of chaige to anyone who has access to the Inteinet anu is
equippeu with a computei oi othei uevice. It can be a fast anu efficient way to ielay
vital infoimation in times of ciisis. Twittei allows foi two-way symmetiical
communication, a two-way conveisation wheie connections aie foigeu anu
feeuback can be given anu ieceiveu (uiunig, 199u, p. 21).
At the time of the 2u1u oil spill, many companies anu oiganizations weie
opeiating at some level on social meuia. Whethei those company leaueis weie
awaie of it oi not, infoimation about the company coulu likely be founu somewheie
on social meuia platfoims. BP, piioi to the oil spill, hau a ielatively inactive Twittei
account . That is, they uiu not use Twittei to actively engage with the public until
aftei the oil spill. In fact, aftei the spill, Twittei accounts such as ÇBPulobalPR
became populai anu gaineu moie followeis than the veiifieu BP account. It seems
S6
cleai that BP was ieactive iathei than pioactive in theii use of social meuia, which
just fuithei uemonstiates theii lack of ciisis communication plan piepaieuness.
Piioi to the ciisis, Biitish Petioleum was not in the best giaces in the eyes of
many uue to the natuie of the inuustiy in which they opeiate. Big oil uiu not neeu a
massive ciisis to be hateu. 0il plays a laige iole in uomestic politics anu foieign
policy. At the time of the spill oil piices hau been soaiing anu many families anu
inuiviuuals in the 0niteu States weie stiuggling financially as a iesult.
Enviionmentalists anu the incieasing numbei of inuiviuuals who, at the time, weie
beginning to view oil as "uiity" uue to the impact of off-shoie uiilling weie calling
foi big oil anu the goveinment to uevelop anu funu cleanei eneigy even befoie the
spill. All of these factois, combineu with the 2uuS Texas City, Texas, explosion, left
BP in a situation of high ieputational thieat going into the ciisis causeu by the 2u1u
explosion anu spill (Coombs, 2uu4).
Consiueiing theii high ieputational thieat at the time of the ciisis, BP shoulu
have auopteu "stiategies that inuicate a gieatei acceptance of iesponsibility foi the
ciisis anu simultaneously uemonstiate concein foi victims" (Coombs, 2uu4, p. 269),
that is, employing social meuia channels to get its siue of the stoiy out anu cieate
both naiiative anu uialogue with its publics. BP showeu lack of piepaieuness anu
woulu have be able to cieate a uiffeient outcome foi themselves hau they taken the
time to ieseaich theii best options anu piepaie. Bau they iealizeu anu accepteu
theii ieputation anu position befoie the spill anu leain fiom the past, they may have
iecognizeu the impoitance anu potential benefits of taking a moie sinceie path.
Focusing on people, the enviionment, anu uamage to the inuustiy iathei than
S7
science anu extiemely technical uetails of the spill woulu have maue them bettei off.
Auuitionally, moie sinceiity fiom CE0 Tony Baywaiu anu an actual apology anu
aumittance of eiioi may have appeaseu fai moie people than the ieactive stiategy
that focuseu on technicalities aftei the spill.
The finuings of this stuuy aie consistent with pievious liteiatuie anu stuuies
about best piactices in social meuia anu best piactices in ciisis communications. BP
uiu not use piopei ciisis communication, anu that was uemonstiateu in theii
seemingly caieless iesponse. They employeu social meuia as a 9+(01"O+ methou anu
uiu not effectively engage anu connect with those using social meuia to ieach out
anu engage with them. They focuseu on veiy technical coiiective actions that they
weie taking to cap the well, stop the flow of oil, anu clean up the mess. 0ften times,
the language that was useu in Tweets was fai too technical foi the aveiage peison
who is not in the inuustiy to unueistanu. Foi example, a Tweet fiom }uly 1S, 2u1u
says "At 1u:Suam we closeu the kill line. At that point the only flow was going out
the choke line." Technical language like that can be uncleai to an aveiage peison
unless they hau knowleuge of what a kill like anu choke line aie. At times uuiing the
spill, BP even useu social meuia to makes jokes about the situation, incluuing one
fiom Nay 27, 2u1u about clean up iesuming as soon as a XXL wet suit was founu foi
a laigei inuiviuual iesponsible foi clean up — a stiategy that cleaily shoulu have
been avoiueu if they weie conceineu about theii image in any way.
Also in accoiuance with pievious liteiatuie was this stuuy's finuings about
the way inuiviuuals useu social meuia to ieach anu connect with BP. The gieatest
peicentage (6S.4%) of Tweets uiiecteu at ÇBP_Ameiica in this stuuy was negative.
S8
Consiueiing the giavity of the situation anu that most people have been founu to
expiess a stiong feeling in eithei uiiection when auuiessing a company oi piouuct
on social meuia, the laige negative social meuia iesponse was to be expecteu. The
laige peicentage of negativity can also be attiibuteu to be ieactive to BP's use of
Twittei uuiing the ciisis. Theii lack of willingness to engage with social meuia useis,
in auuition to theii iesponse stiategies, woikeu against them as eviuenceu by the
finuings in this stuuy.
%):9=
Social meuia has changeu the way inuiviuuals communicate. Two-way
communication thiough methous othei than meeting face-to-face oi thiough
telephones is now the noim. Communication platfoims like Twittei anu othei social
meuia have biiugeu a gap between inuiviuuals anu companies. The neeu foi
tianspaiency, honesty, anu connection is gieatei than evei because of how
accessible eveiyone has become. It is no longei necessaiy to wiite a lettei oi place a
phone call to voice youi opinion about a company oi situation. Piess ieleases, piess
confeiences, anu tiauitional methous of public ielations aie now just pait of the
piactice of builuing anu maintaining ielationships with the public because of social
meuia.
Buiing times of ciisis, social meuia has become a vital outlet to gathei anu
senu infoimation. It has become expecteu foi an oiganization, inuiviuual, oi
company to use social meuia to auuiess conceins anu uisseminate infoimation
uuiing times of ciisis whethei they aie emeigency ciises, coipoiate ciises, technical
malfunctions, oi otheiwise. BP's slow ieaction on social meuia anu unwillingness to
S9
engage with conceineu citizens looking to engage on Twittei anu othei social meuia
was just fuel to the fiie uuiing the oil spill anu its clean up.
Although the well has been cappeu in the uulf anu billions of uollais have
been spent on the clean up anu on compensation measuies foi citizens in the uulf,
BP still faces haish ciiticism foi theii hanuling of the oil spill. As time passes anu
moie infoimation comes out about BP piactices befoie anu uuiing the spill, in
auuition to the amount of uamage uone to the enviionment, inuustiy, anu health of
uulf citizens, BP has to continue to auuiess the ciisis anu piactice ciisis
communication. Theii use of social meuia has evolveu fiom being neaily non-
existent to being active on many platfoims anu also uemonstiating slightly moie
willingness to engage.
4:&#"0(1"7,*
Theie aie seveial implications that iesult fiom this stuuy. The fiist is that the
stuuy uemonstiates the neeu foi companies anu oiganizations to have a ciisis
communication plan in place that caiefully selects the stiategies that they will use in
accoiuance with best piactices. Some ieseaich woulu have given eviuence to
suppoit that BP shoulu have been less technical anu moie apologetic anu sinceie in
the wake of the uisastei. Some foiethought about using social meuia in the ciisis
communication plan woulu have alloweu them to be moie pioactive iathei than
ieactive anu moie engaging iathei than using the platfoim to uistiibute technical
infoimation. Ciisis communication has uemonstiateu in the past how it can be useu
to tiansfoim situations anu allow embattleu oiganizations to oveicome anu move
on aftei the ciisis enus. BP uiu not have to look to fai to see what woiks anu what
4u
uoes not woik in ciisis communication. Although the situation was unique in many
ways, the inuiviuuals iesponsible foi the ciisis communication plan coulu have
lookeu to the past. Specifically, they coulu have avoiueu Exxon's mistakes.
Anothei implication fiom this stuuy foi public ielations piofessionals
incluues knowing the impoitance of knowing youi auuience anu theii conceins. Bau
BP known theii auuience anu gaugeu theii conceins, they coulu have uiiectly
auuiesseu them. Social meuia allows foi monitoiing what people aie saying about a
paiticulai company thiough channels such as blogs anu Tweets. Taking the time to
seaich social meuia anu engage those with conceins, ciiticisms, oi even suppoit
helps bettei iuentify conceins anu auuiess them. BP's Twittei account uiu not
miiioi the conceins of those tiying to connect with them. The social meuia in
geneial, anu Twittei in paiticulai, aie moie than platfoims foi enteitainment; they
have uemonstiateu theii seiious implications anu powei anu shoulu not have been
oveilookeu anu misuseu as they weie by BP. BP's negligent behavioi was calleu to
attention in a Tweet fiom ÇNackcolliei who saiu "Seiiously if ÇBP_Ameiica uoesn't
caie enough to monitoi Twittei oi engage heie, what makes us think they caie
about ÇBPulobalPR.".
This stuuy showeu how inuiviuuals useu social meuia in times of ciisis to
voice theii opinions. In this case, theie was no shoitage of Tweets haishly ciiticizing
eveiy move that BP was making. 0n Nay S, 2u1u ÇNatuialPapa Tweeteu "Bey
ÇBP_Ameiica - why aie citizens volunteeiing to clean youi fieakin' mess. Put those
piofits to goou use anu clean up aftei youiself." Fuithei a Tweet fiom }une 4, 2u1u
geneiateu oiiginally fiom ÇLuxuiyPRual but ie-Tweeteu numeious times stateu
41
"BP is spenuing $Su NILLI0N on a new au campaign. They shoulu spenu that on
cleaning up the ocean. RT if u agiee ÇBP_Ameiica." Stiategic planning anu piopei
employment of social meuia in the futuie can allow foi public ielations piactitioneis
to achieve bettei outcomes than the outcomes ieacheu by the BP public ielations
plan.
2":"1(1"7,* (,- ?"9+01"7, /79 U)91>+9 I+*+(90>
This stuuy focuseu only on one aspect of social meuia, Twittei, anu its use in
ciisis communications. The iesults of this stuuy uemonstiate how BP's use of Image
Restoiation stiategy was useu in this one singulai outlet, anu how the public
ieacteu to the situation on just that one outlet. While social meuia is veiy impoitant,
it was just one aspect of BP's ciisis iesponse anu it was only one way that the public
ieacteu aftei the Apiil 2u1u uulf 0il Spill. Twittei is a populai way to vent
fiustiations anu feelings, but the foimat of the content uoes make it veiy uifficult to
inteipiet feeling as some things like saicasm anu humoi aie not easily unueistoou
in ceitain contexts.
Fuithei ieseaich shoulu incluue othei social meuia outlets to ueteimine if
theie was any vaiiance between Image Restoiation Stiategies useu acioss the
platfoims anu also to ueteimine if social meuia useis tone vaiieu acioss the
uiffeient platfoims in the messages uiiecteu towaius BP. Noie ieseaich may also
incluue iuentifying any uiffeience in tones anu message ieception in uiffeient
geogiaphic aieas. This stuuy uoes not take into consiueiation wheie the message
senuei lives anu whethei oi not they weie uiiectly affecteu by the spill oi not.
42
Fuithei ieseaich in these paiticulai aieas coulu iuentify vaiiances anu allow foi
bettei, moie thoiough ciisis communication planning in the futuie.
4S
Refeiences
Auei, N.R. (2u11). The policy sciences of social meuia. M>+ 87#"0= %1)-"+* V7)9,(#5
TWXYZ[ 7u9 - 7S2.
Benoit, W.L. (1997). Image iepaii uiscouise anu ciisis communication. 8)D#"0
I+#(1"7,* I+O"+@5 ST\S[ 177 - 184.
Biaufoiu, }.L. & uaiiett, B.E. (199S). The effectiveness of coipoiate communicative
iesponses to accusations of unethical behavioi. V7)9,(# 7/ G)*",+** .1>"0*5 QY5
87S - 892.
Biitish Petioleum. (2u1u). BP confiims that Tiansocean Ltu issueu the following
statement touay (Piess Release). Retiieveu fiom http:¡¡bp.com.
Biitish Petioleum. (2u1u). BP offeis full suppoit to Tiansocean aftei uiilling iig fiie
(Piess Release). Retiieveu fiom http:¡¡www.bp.com.
Biitish Petioleum. (2u1u). BP offeis sympathy to the families of those lost in the 0S
oil iig fiie (Piess Release). Retiieveu fiom http:¡¡www.bp.com.
Cheong, N. & Lee, v.C.S. (2u11). A micioblogging-baseu appioach to teiioiism
infoimatics: Exploiation anu chionicling civilian sentiment anu iesponse to
teiioiism events via Twittei. 4,/791"7, %=*1+: U97,15 QT[ 4S - S9.
Chess, C. (2u1u). BP's iisk communication uisastei. .)97&+(, V7)9,(# 7/ I"*6
I+3)#(1"7,5 T:211 - 217.
Coombs, T.W. (2uu7). Attiibution theoiy as a guiue foi post-ciisis communication
ieseaich. 8)D#"0 I+#(1"7,* I+O"+@5 TT[ 1SS - 1S9.
Coombs, T.W. (2uu4). Impact of past ciises on cuiient ciisis communication:
Insights fiom situational ciisis communication theoiy. V7)9,(# 7/ G)*",+**
'7:,"0(1"7,5 YQ[ 26S - 289.
Buuley, B. (2u1u, Nay). G7D ?)-#+= (1 1>+ 0>"+/ +E+0)1"O+*H 0#)D 7/ G7*17,\ Speech
piesenteu at the Chief Executives' Club of Boston, Boston, Na.
Euosomwan, S., Piakasan, S.K., Kouame, B., Watson, }., & Seymoui, T. M>+ V7)9,(# 7/
N&&#"+- <(,(3+:+,1 (,- .,19+&9+,+)9*>"&5 S]QQ5 Q^[ 79 - 91.
Fischei, E. & Reubei, R.A. (2u11). Social inteiaction via new social meuia: (Bow) can
inteiactions on Twittei affect effectual thinking anu behavioi. M>+ V7)9,(# 7/
G)*",+** F+,1)9",35 S^[ 1 - 18.
uiattan, L. & Reese, A. (2u11). Bepiession: Bioauei psychological impact of 2u1u BP
oil spill. .,+93= K++6#= L+@*5 <(90> Y5 S]QQ[ S8S - S86.
44
uiiggs, }.W. (2u11). BP gulf of mexico oil spill. .,+93= G(9 N**70"(1"7, .,+93= 2(@
V7)9,(#5 TS: 1 - 12.
uiunig, }.E. (199u). Theoiy anu piactice of inteiactive meuia ielations. 8)D#"0
I+#(1"7,* P)(91+9#=5 T_[ 18 - 2S.
Bailow, W.F., Biantley, B.C., & Bailow, R. (2u11). BP initial image iepaii
stiategiesaftei the Beepwatei Boiizon spill. 8)D#"0 I+#(1"7,* I+O"+@5 T`[ 8u -
8S.
Bolsti, 0.R. (1969). '7,1+,1 (,(#=*"* /79 1>+ *70"(# *0"+,0+ (,- >),"1"+*\ Reauing,
NA.: Auuison - Wesley Publishing Company.
}ansen, B.}., Zhang, N., Sobel, K. & Chowuuiy, A. (2uu9). Twittei powei: Tweets as
electionic woiu of mouth. V7)9,(# 7/ 1>+ N:+9"0(, %70"+1= /79 4,/791"7, (,-
M+0>,7#73=5 ^]XQQZ[ 2169 - 2188.
NcCauley, L. (2u1u). Enviionments anu health: Will the BP oil spill affect oui health.
N:+9"0(, V7)9,(# 7/ L)9*",35 QQ][ S4 - S6.
NcSwite, 0.C. (2uu9). The challenge of social netwoiks. N-:","*19(1"O+ M>+79= (,-
89(E"*5 TQXQZ[ 78 - 9S.
Noiozov, E. (2u1u). Cognitive suiplus: Cieativity anu geneiosity in a connecteu age.
G7*17, I+O"+@5 T_[ 24 - 26.
Nuialiuhaian, S., Billistone, K., & Shin, }. (2u11). The gulf coast oil spill: Extenuing
the theoiy of image iestoiation uiscouise to the iealm of social meuia anu
beyonu petioleum. 8)D#"0 I+#(1"7,* I+O"+@5 T`: 226 - 2S2.
Ritchie, L.A., uill, B.A., & Picou, }.S. (2u11). The BP uisastei anu an Exxon valuez
ieiun. N:+9"0(, %70"7#73"0(# N**70"(1"7,5 %)::+9 S]QQ[ Su - SS.
Smith, L.C., Smith, L.N., & Ashcioft, P.A. (2u1u). Big oil, big consequences, anu the
big unknown: Exploiing the legal, iegulatoiy, anu enviionmental impact of
the uulf oil spill. N#D(,= 2(@ I+O"+@5 `Y[ 1 - 24.
Stacks, B.W., & Nichaelson, B. (2u1u). N &9(01"1"7,+9H* 3)"-+ 17 &)D#"0 9+#(1"7,*
9+*+(90>5 :+(*)9+:+,15 (,- +O(#)(1"7,. New Yoik: Business Expeit Piess.
Wateis, R.B., Tinuall, N.T.}., & Noiton, T.S. (2u1u). Neuia catching anu the jouinalist-
public ielations piactitionei ielationship: Bow social meuia aie changing the
piactice of meuia ielations. V7)9,(# 7/ 8)D#"0 I+#(1"7,* I+*+(90>5 SSXTZ: 241 -
261.
4S
Weinei, B. (198S). A), :71"O(1"7,\ New Yoik: Spiingei
46
APPENBIX A
;>,8&8 A):&*0
B*&$8 )- #*/618&8C
The stuuy examineu social meuia messages by Biitish Petioleum (BP) anu otheis.
Seveial units of analysis will be examineu in iesponse to the social meuia content
geneiateu by BP Ameiica as a iesponse to the oil spill.
.)=&/6 ?,:&/ ;15,C
D! Comments geneiateu by the public towaiu ÇBP_Ameiica on Twittei
E! Twittei content geneiateu by ÇBP_Ameiica
7,8$)%/$&)* .$%/$,01C
D! Restoiation stiategy
E! Not iestoiation stiategy
@4/0, 7,8$)%/$&)* .$%/$,01 FG,*)&$H DIIJKC
D! .&456, +,*&/6 - uefineu as when the accuseu paity iesponus to any
accusations by iefuting claims of iesponsibility
E! .>&-$&*0 G6/4, - uefineu as when the accuseu paity iesponus to
accusations by placing culpability on anothei paity that may oi may not
have been involveu
L! M%)N)=/$&)* - uefineu as when the accuseu paity aumits they aie to
blame, but that they weie piovokeu into committing the haimful action
O! +,-,/8&'&6&$1 - uefineu as when the accuseu paity iesponus to
accusations by saying that they lackeu contiol oi infoimation neeueu to
pievent the actions they aie accuseu of
47
P! #==&:,*$8 - uefineu as when the accuseu paity iesponus to accusations
by stating that the event was unintentional
Q! 9)): @*$,*$&)* - uefineu as when the accuseu paity iesponus to
accusations by explaining that the event was the iesult something
planneu with goou puipose that went wiong inauveitently
J! G)68$,%&*0 - uefineu as when the accuseu paity highlights positive
chaiacteiistics they holu which they believe outweigh the bau in the
situation
R! ?&*&4&S/$&)* - uefineu as when the accuseu paity uownplays the
seiiousness oi giavity of the situation
I! +&--,%,*$&/$&)* - uefineu as when the accuseu paity iesponus to the
accusations by explaining that the event is uiffeient fiom othei acts anu
cannot be compaieu to oi unueistoou thiough othei events
DT! ;%/*8=,*:,*=, - uefineu as when the accuseu paity iesponus to
accusations by explaining that the event shoulu be unueistoou in a
uiffeient context than it is cuiiently being unueistoou
DD! #$$/=U #==(8,% - uefineu as when the accuseu paity tuins the blame
aiounu anu puts it on the paity that is accusing them of causing the
pioblem
DE! A)45,*8/$&)* - uefineu as when the accuseu paity iesponus to the
accusations by stating that those impacteu will be able to iecovei
uamages, but uoes not necessaiily mean that the accuseu is takings
iesponsibility foi causing the uamage
48
DL! A)%%,=$&N, #=$&)* - uefineu as when the accuseu paity takes
iesponsibility foi unueitaking necessaiy measuies to fix the pioblem,
but uoes not necessaiily take iesponsibility foi causing the pioblem
DO! ?)%$&-&=/$&)* - uefineu as when the accuseu paity aumits iesponsibility
foi a ceitain event anu asks foigiveness fiom the geneial public anu
those impacteu
*,C
D! M)8&$&N,÷favois the actions taken ¡ comments maue by BP
E! Ne0ative÷ciiticizes the actions taken ¡ comments maue by BP
L! Neutial÷neithei favois noi ciiticizes the actions taken ¡ comments maue
by BP oi theie aie a seemingly equal amount of positive anu negative
comments
9,*:,% )- 8,*:,%C
D! Female
E! Nale
L! 0thei
49
?,88/0, <)=(8C
D! Inuustiy
E! Bealth
L! Enviionment
O! Financial anu economic impact
P! Recoveiy
Q! 0thei
+/$, )- ?,88/0,
@*8$%(=$&)*8C
Coueis aie iesponsible foi accuiately categoiizing all uata gatheieu as
coiiectly as possible. Fiist, coueis enteieu the uate that the unit of analysis was
wiitten in a NN¡BB¡YY foimat. Seconu, when applicable, coueis categoiizeu the
wiitei of the unit being analyzeu as Nale, Female, oi 0thei. Because the content
being analyzeu was coming fiom social meuia outlets, theie may be some names
that aie aliases oi not ieal names. These weie categoiizeu unuei "0thei," as well as
any names that aie unisex oi otheiwise not cleai. Names that aie tiauitionally
women's names weie categoiizeu as "Female" anu tiauitionally masculine names as
"Nale."
All uata in the sample geneiateu fiom BP was collecteu fiom the company's
veiifieu Twittei account. These tweets weie coueu foi uate, iestoiation stiategy,
anu when applicable, the paiticulai Image Restoiation Stiategy. To coue content foi
Image Restoiation Stiategy (Benoit, 1997), coueis iuentifieu the stiategy BP
Su
employeu in cieating the message content. Coueis weie given a list of the 14
stiategies anu examples of each befoie the couing piocess begins.
The uata set that is composeu of comments geneiateu by the public uiiecteu
towaius BP was then coueu to iuentify tone, iefeience to paiticulai subject aiea,
uate, anu genuei of wiitei. Content that was wiitten in suppoit of BP oi theii effoits
to clean the uulf, iectify the situation, oi take iesponsibility foi what they have uone
was categoiizeu as "Positive." Also to be categoiizeu as positive weie comments
about measuies that aie peiceiveu as effective oi comments about the
impiovements that BP is iesponsible foi in the wake of the oil spill.
Content that conuemneu BP's posteu message, theii measuies in the uulf,
anu theii iesponse to the oil spill in geneial weie categoiizeu as "Negative." Content
that ciiticizeu BP's iesponse as ineffective, insufficient, oi impiopei was categoiizeu
as negative. Content that challenges BP's honesty anu integiity in the cleanup oi
afteimath in any way was also categoiizeu as negative.
Some content hau both positive anu negative iemaiks towaiu BP anu theii
iesponse to the oil spill. Content that incluueu equal paits negative anu positive was
categoiizeu as "Neutial." Also to be categoiizeu as neutial, weie those that geneially
expiesseu no stiong feeling in eithei uiiection about the actions being unueitaken.
Foi example, some iesponses might give a suggestion oi claiification, but weie
neithei ciitical noi wiitten in piaise. This content was categoiizeu as neutial.
The content in this sample set was coueu to ueteimine the paiticulai focus of
the message, incluuing: inuustiy, health, enviionment, financial anu economic
impact, iecoveiy, oi othei.
S1
Befoie the couing anu categoiization piocess began, coueis uiscusseu anu
came to agieement on the meaning of the categoiies above. 0nce agieement anu
unueistanuing was ieacheu, couing begin. Inteicouei ieliability will be testeu to
ensuie that a high level of agieement is ieacheu.
S2
APPENBIX B
Scieen shot of Tweets fiom ÇBP_Ameiica
SS
APPENBIX C
Scieen shot of Tweets ÇBP_Ameiica
doc_434314405.pdf
Scholarly Repository
Open Access Teses Electronic Teses and Dissertations
2012-05-08
British Petroleum's Use of Image Restoration
Strategy on Social Media and Response Afer the
2010 Gulf Oil Spill
Katherine Rogers
University of Miami, [email protected]
Follow this and additional works at: htp://scholarlyrepository.miami.edu/oa_theses
Tis Open access is brought to you for free and open access by the Electronic Teses and Dissertations at Scholarly Repository. It has been accepted for
inclusion in Open Access Teses by an authorized administrator of Scholarly Repository. For more information, please contact
[email protected].
Recommended Citation
Rogers, Katherine, "British Petroleum's Use of Image Restoration Strategy on Social Media and Response Afer the 2010 Gulf Oil
Spill" (2012). Open Access Teses. Paper 334.
UNIVERSITY OF MIAMI
BRITISH PETROLEUM’S USE OF IMAGE RESTORATION STRATEGY ON
SOCIAL MEDIA AND RESPONSE AFTER THE 2010 GULF OIL SPILL
By
Katherine Rogers
A THESIS
Submitted to the Faculty
of the University of Miami
in partial fulfillment of the requirements for
the degree of Master of Arts
Coral Gables, Florida
May 2012
©2012
Katherine Rogers
All Rights Reserved
UNIVERSITY OF MIAMI
A thesis submitted in partial fulfillment of
the requirements for the degree of
Master of Arts
BRITISH PETROLEUM’S USE OF IMAGE RESTORATION STRATEGY ON
SOCIAL MEDIA AND RESPONSE AFTER THE 2010 GULF OIL SPILL
Katherine Rogers
Approved:
________________ _________________
Don W. Stacks, Ph.D. Shannon Campbell, Ph.D.
Professor of Public Relations Assistant Professor of Public
Relations
________________ ________________
Joseph Treaster, M.S. Terri A. Scandura, Ph. D.
Professor, John S. and James L. Knight Chair Dean of Graduate School
in Cross-Cultural Communication
ROGERS, KATHERINE (M.A., Communications)
British Petroleum’s Use of Image Restoration (May 2012)
Strategy on Social Media and Response After the
2010 Gulf Oil Spill
Abstract of a thesis at the University of Miami.
Thesis supervised by Professor Don Stacks.
No. of pages in text. (53)
As the population and use of social media platforms increase, so does the role
they play in crisis communication. This study focuses on how BP used social media,
specifically Twitter, as a platform for their public relations response to the devastating
2010 Gulf Oil Spill. Additionally, this study focuses on identifying how social media
users used Twitter as an avenue to reach out to BP on and voice concerns and attitudes
after the spill. A content analysis of Tweets written in response to the spill was conducted
and through that analysis it was determined that BP used Twitter as a way to disperse
messages about how they were making things better or compensating people; while social
media users used Twitter to attempt to engage with BP and voice negative opinions about
many areas that were damaged because of the spill. By identifying how BP used Twitter
and restoration strategies in the wake of the spill, it is possible to compare their use to
best practices that have developed through academic research and existing case studies
where social media and restoration strategy was used effectively. This can identify why
BP’s response was not well received. By identifying how social media users used
Twitter, it is possible to use this case study as an example of the importance of engaging
with users and identifying areas of concern. This study has implications for public
relations practitioners and all professionals who are charged with creating crisis
communication plans. Limitations of the study and recommendations for future research
were included.
iii
!"#$% '( )'*!%*!+
PAuE
LIST 0F TABLES .................................. iv
CBAPTER
1 INTR0B0CTI0N ............................ 1
2 LITERAT0RE REvIEW ........................ 4
0il Spill Cause anu Effects ......................... S
Legal Issues, Lack of Piepaieuness anu Impact on Clean-up .... 6
Similaiities between Exxon valuez anu BP Spill .......... 7
BP's Safety Recoiu anu 0se of Social Neuia .............. 8
Impact of 0il Spill on the Bealth of uulf Coast Citizens ........ 1u
Two Theoietical Ciisis Nouels ..................... 1u
Attiibution Theoiy anu Situation Ciisis Communication Theoiy . 11
Ciisis Communication Response Nouels .............. 1S
Coipoiate Communicative Response Nouel ............ 1S
A Changing Neuia Enviionment .................... 17
Social Neuia Revolution anu Ciisis Communication ........ 17
The Enviionmental Revolution ................... 2u
Reseaich Questions .......................... 21
S NETB0B0L0uY ............................ 22
0nits of Analysis ........................... 22
Sampling ................................ 22
Couing .................................. 2S
Couing Reliability ............................ 24
Statistical Analysis .......................... 24
4 RES0LTS ................................ 2S
Besciiptive Statistics .......................... 2S
Test of Reseaich Questions ...................... 27
S BISC0SSI0N ............................... S4
Summaiy ............................... S8
Implications ............................... S9
Limitations anu Biiection foi Fuithei Reseaich .......... 41
REFERENCES ................................... 4S
Appenuix A .................................. 46
Appenuix B .................................. S2
Appenuix C .................................. SS
iv
$,+! '( !"#$%+
Table 2-1: Attiibution Theoiy ........................... 12
Table 2-2: Benoit's (1997) Image Restoiation Stiategies ............. 14
Table 2-S: Biaufoiu anu uaiiet's (199S) Response Nouel ............. 16
Table 4-1: uenuei of Tweet Senuei ........................ 26
Table 4-2: Peicentage of Tweets Containing Restoiation Stiategy .......... 26
Table 4-S: Recoiueu Tone of Tweets Biiecteu Towaius BP ........... 28
Table 4-4: Recoiueu Focuses of Nessages ..................... 29
Table 4-S: Tone - Focus Cioss Tabulation ..................... Su
Table 4-6: uenuei - Tone Cioss Tabulation ................... Su
Table 4-7: Focus - Tone Cioss Tabulation ...................... Su
Table 4-8: Image Restoiation Stiategy Fiequencies ............... SS
1
Chaptei 1
Intiouuction
The oil spill in the uulf of Nexico that began on Apiil 2u, 2u1u, was the iesult
of an explosion anu fiie ioughly 4u miles off the coast of Louisiana. Biitish
Petioleum (BP) owneu anu opeiateu the Beepwatei Boiizon uiilling unit woiking
on an off shoie exploiatoiy well when malfunctions causeu the explosion. The
explosion claimeu the lives of 11 men woiking on the exploiatoiy well. As the fiie
buineu, oil pouieu fiom the well into the uulf of Nexico at an estimateu iate of
SS,uuu baiiels of oil pei uay. This continueu until }uly 1S, 2u1u, when the well was
finally cappeu. In total, between 18S anu 2uS million gallons of ciuue oil pouieu into
the uulf. In auuition to the tiagic loss of life, the spill pioveu to be uevastating to
maiine life in the uulf, killing anu enuangeiing countless animals anu entiie species.
uulf Coast inuustiies such as fishing, touiism, anu ieal estate iecoiueu uamages in
the billions of uollais. The spill took a physical anu psychological toll on those who
weie impacteu by its afteimath, both uiiectly anu inuiiectly.
Total uamages to BP, the enviionment, anu the economy in the uulf Coast aie
cuiiently estimateu to be $S6.9 billion. It was a uisastei on both human anu
enviionmental fionts anu Biitish Petioleum was left to figuie out how to combat all
aspects of the situation. BP nevei took uiiect iesponsibility foi the spill, iathei in
piess ieleases anu othei communications issueu following the explosion anu spill,
they focuseu on shifting blame anu assuiing the public they woulu honoi all
legitimate claims anu gave uetaileu infoimation about the steps that they weie
taking to stop the flow of oil anu iectify the situation. BP was faceu with a majoi
2
public ielations ciisis in the wake of this uisastei. They encounteieu many
challenges anu setbacks, incluuing insensitive iemaiks maue by CE0 Tony Baywaiu
anu complaints fiom the public that BP was not uoing enough iemeuy the situation.
The BP Beepwatei Boiizon 0il Spill began at the height of the enviionmental
movement. People weie incieasingly vigilant of enviionmental issues anu many
weie beginning to make a conscious effoit to become moie enviionmentally
iesponsible oi "gieen." Coipoiate social iesponsibility was becoming of gieat
impoitance anu consumeis consiueieu moie than just the piouuct anu piice when
making a puichase.
Piioi to the spill, oil itself was a contioveisial issue. uas piices weie soaiing
anu many Ameiicans weie taking a seiious financial hit because of the inciease.
With that being saiu, the issue of off shoie uiilling was becoming incieasingly
contioveisial even befoie the spill uue to conflicting conceins about the
enviionment anu iising cost of oil anu the iesulting incieaseu piice of gas. 0ff shoie
uiilling anu enviionmental policies playeu a pivotal iole in the 2uu8 piesiuential
election. Aftei the spill, Piesiuent Baiak 0bama hau to put in place ceitain
emeigency measuies anu policies, incluuing a moiatoiium on off shoie uiilling that
lasteu foi seveial months befoie being lifteu.
Fuithei, the spill occuiieu when social meuia was becoming incieuibility
impoitant anu influential in the meuia anu society. Euosomwan, Piakasan, Kouame,
Watson, anu Seymoui (2u11) state that "social meuia helps conveisations to ieach a
wiuei auuience leveiaging the 'long tail' concept, which means conveisations that
can be conveyeu to uiffeient foiums" (p. 8S). Social meuia platfoims staiteu
S
incieasing in ielevance in 2uuu. By }uly 2u1u, Facebook hau moie than Suu million
active useis, YouTube viueos weie ieceiving millions of views uaily, anu Twittei was
becoming incieasingly populai as a souice foi news anu communication
(Euosomwan et. al, 2u1u, pp. 81 - 84). News was tiaveling fastei thanks to social
meuia, anu anyone's opinion coulu be wiitten anu voiceu. Nany coipoiations anu
companies weie opeiating social meuia platfoims anu hau to have a plan foi theii
social meuia use in times of ciisis.
The BP oil spill anu its afteimath pioviue valuable lessons in ciisis
communication. It is a case that will seive as an example of what to uo anu what
shoulu be avoiueu when piepaiing a ciisis communication plan oi combating a
ciisis. The long teim effects of the spill aie not yet known, but in the shoit teim
theie aie lessons to be leaineu about not only what coulu have been uone to pievent
the spill but about how to combat an enviionmental ciisis on the public ielations
fiont. The puipose of this stuuy is to iuentify, specifically, the ciisis iesponse
stiategies employeu by Biitish Petioleum on Twittei anu why they may not have
been optimally effective in using social meuia to contiol the public ielations ciisis
that iesulteu fiom the explosion anu oil spill. Auuitionally, it will examine in uepth
how people iesponueu anu ieacteu on social meuia outlets.
4
Chaptei 2
Liteiatuie Review
In oiuei to examine the iole of ciisis communication stiategy anu the use of
social meuia as a platfoim to expiess opinion anu attituue in the BP 0il Spill
iesponse anu afteimath, theie aie seveial aieas of influence to be ievieweu.
0nueistanuing exactly how the oil spill was causeu anu the extent of the uamages
that iesulteu, incluuing uamages to the enviionment, economy, anu citizens' health
anu well-being pioviues eviuence as to why people weie so uismayeu by the spill
anu possessing the knowleuge that it coulu have been pieventeu with some basic
safety piecautions.
Legal issues weie plentiful aftei the spill. Questions of piepaieuness, guilt,
liability, anu iights of those effecteu abounueu aftei the uisastei. BP coulu have
lookeu to pievious enviionmental uisasteis anu how they weie hanuleu foi
guiuance as they moveu foiwaiu. The 1989 Exxon valuez oil spill in paiticulai
pioviueu a pieceuent as to how oil spills can be hanuleu in the 0niteu States. Ciisis
communication anu image iestoiation mouels (e.g., Benoit, 1997) aie in existence to
help oiganizations anu inuiviuuals piepaie foi anu iecovei fiom a ciisis. In auuition,
touay the social meuia must be auuiesseu in ciisis communication. Social meuia has
cieateu a sense of tianspaiency anu accessibility between citizens anu
oiganizations. A plan must auuiess the way it will be employeu in ciisis
management situations.
Finally, cuiient social tienus neeu to be taken into consiueiation uuiing
ciisis communication planning. The histoiy anu state of the "gieen" movement in
S
the 0niteu States anu the ability to know how it will impact ciisis management is
impoitant in the BP oil spill case because of the enviionmental uestiuction it causeu.
0il uiilling is a contioveisial enviionmental issue anu the spotlight on
enviionmental issues at the time of the spill maue the fallout even woise foi BP.
!"# %&"## '()*+ (,- .//+01*
An explosion on the Beepwatei Boiizon uiilling unit being opeiateu by
Biitish Petioleum ioughly 4u miles off the coast of Louisiana tiiggeieu the oil spill
that began on Apiil 2u, 2u1u, in the uulf of Nexico. The explosion was just the
beginning of a nightmaie foi BP, iesiuents of the uulf Coast, anu all those impacteu
by the spill. Issues of iesponsibility, accountability, liability, anu cause abounueu in
the wake of the spill as citizens of the uulf Coast watcheu theii enviionments anu
lives be giavely impacteu by the oil spilling into the uulf. Smith, Smith, anu Ashcioft
(2u1u) explain that the uamages fiom the spill aie
attiibuteu to thiee majoi factois: (1) human eiioi anu equipment failuie at
BP's Beepwatei Boiizon offshoie uiilling unit; (2) failuie of the 0.S.
goveinment to assign, anu in some cases to peimit, iesouices to assist with
the containment of the oil spill; anu (S) misinfoimation uisseminateu by the
news meuia iegaiuing the amount anu location of oil pollution in the watei
anu on the beaches of the uulf of Nexico. (p.1)
The iesults of the spill weie not easy to account foi, howevei; estimateu total
uamages to BP, the enviionment, anu the economy aie at least $S6.9 billion, though
this is subject to change. Some shoit-teim effects of the spill associateu with the
iesilience of the enviionment anu the oil spill aie known now, but the long-teim
effects will not be iealizeu foi yeais to come. The oil spill leu to laws anu iegulations
a moiatoiium on uiilling (which was latei iuleu aibitiaiy anu oveituineu) anu
iestiictions to aieas foi fishing (Smith et. al., 2u1u, p. S)!
6
2+3(# 4**)+*5 2(06 7/ 89+&(9+-,+** (,- 4:&(01 7, '#+(,
& Along with the enviionmental anu humanitaiian uisastei that BP was
iesponsible foi taking caie of, they now hau to ueal with a massive public ielations
ciisis. Eviuence points to the fact that the ciisis coulu have been avoiueu all togethei
if piopei safety piecautions weie taken. Bowevei, the explosion happeneu anu oil
pouieu into the uulf at an alaiming iate. While theii piepaieuness to ueal with the
spill enviionmentally was being questioneu, theie weie also issues with whethei oi
not they hau a piopei ciisis communication plan in place. uiiggs (2u11), foi
example, iaises issues of piepaieuness. 0ne of the majoi issues that came into play
legally uuiing the spill was the pace of iesponse uue to the fact that BP woulu not
allow any piivate help without going thiough the lengthy peimission piocess. This
was a battle fought by local people who weie familiai with the aiea anu equippeu to
help in some ways, but weie piohibiteu fiom uoing so. This was ieflecteu in the
meuia as a souice of gieat fiustiation. At the ioot of this pioblem was the legal issue
that BP was financially iesponsible foi the cleanup anu theiefoie hau to appiove
eveiything. If they alloweu foi outsiue help befoie appioval they iiskeu the piivate
paity wanting to collect foi the woik they hau uone (uiiggs, 2u11, p. 7).
BP was not piepaieu foi the PR ciisis they woulu be uealing with the in the
afteimath of the spill. The Society foi Risk Analysis (SRA) <",";*=:&7*"): 7, 1>+
?++&@(1+9 A79"B7, !"# %&"## (2u1u) biought to question BP's piepaieuness on
seveial levels. In paiticulai, past SRA piesiuent Caion Chess questioneu BP's level
of iisk communication piepaieuness. Chess felt that BP lackeu any amount of
piepaieuness anu faileu miseiably fiom a iisk communications point of view.
7
Fuithei, Chess noteu similaiities between BP's iesponse anu Exxon's 1989 iesponse
to its oil spill in the Stiaits of valuez, Alaska. She citeu an aiticle fiom C9++,@"9+ that
noteu "uepiessing similaiities between Exxon's anu BP's iisk communication
smash-ups: shifting blame fiom the coipoiation, emphasizing science iathei than
people, anu initially uownplaying the potential impact of the uisastei" (Chess, 2u1u,
p. 216).
%":"#(9"1"+* D+1@++, .EE7, F(#-+B %&"## (,- G8 %&"##
BP was not without an example on how to ueal with spills. They hau plenty of
lessons that they coulu have taken away fiom the 1989 Exxon valuez oil spill. It is
impossible to uiscuss the iesponse to the BP oil spill without looking at othei majoi
inciuents that pieceueu it. The Exxon valuez oil spill seives as benchmaik of soits to
compaie the BP iesponse to, although the BP spill was laigei than the Exxon spill.
Ritchie et al. (2u11) note that one of the lessons leaineu fiom the Exxon spill is that
the BP spill iecoveiy is fai fiom ovei:
We know that social anu psychological stiess is heighteneu by the
unceitainty that comes fiom toxic contamination. The complexities of local
anu iegional economies anu the upheaval cieateu by a tempoiaiy economic
boom cieateu by cleanup opeiations also contiibute to the uisastei's
lingeiing effects. Finally, as injuieu paities seek compensation foi uamages
iesulting fiom this enviionmental uisastei, many will be tiaumatizeu by
impeisonal buieauciatic stiuctuies anu piotiacteu litigation. (p. S1)
Anothei similaiity is that both spills effecteu ienewable iesouice communities;
ienewable iesouice communities "uepenu on ienewable natuial iesouices foi theii
social, cultuial anu economic existence" (p. S1). Ritchie also uiaws similaiities
between those in the aiea of the spill who gaineu financially as a iesult of the spill
because of the cleanup effoits. Fuithei, uisasteis often uiive people to move fiom
8
theii communities. This iesults in economic loss anu human capital loss (Ritchie,
2u11, p. S4).
G8H* %(/+1= I+079- (,- J*+ 7/ %70"(# <+-"(
Buiing times of ciisis, a company's safety iecoiu anu image come into
question. Companies who have bettei ieputations going into a ciisis often come out
bettei than those who go into the ciisis unfavoiably. BP is no exception. The oil spill
has biought ciiticism fiom a wiue iange of people, fiom activist gioups to
goveinment entities. While BP has taken iesponsibility foi taking caie of the spill,
they have shifteu blame on subcontiactois Tiansocean anu Ballibuiton. A stuuy by
Nuialiuhaian, Billistone, anu Shin (2u11) choose to examine social meuia anu how
BP useu it uuiing the oil spill iesponse because it allows them to unueistanu "the
uimensions of auuience ieaction" which they claim to be "ciucial in a cycle of ciisis"
(p. 227). The social meuia seiveu as a platfoim foi the public to voice theii
conceins, issues, anu angei uuiing the spill. Nany times, the messages weie in
iegaius to BP's safety measuies in the past leauing up the spill anu how the spill
coulu have been pieventeu hau appiopiiate safety measuies been taken.
BP's safety iecoiu came into question initially in the eaily 2uuus. They came
unuei fiie even moie so aftei a 2uuS explosion at its Texas City, Texas, iefineiy that
claimeu the lives of 1S woikeis anu injuieu neaily 2uu otheis. Aftei the explosion,
BP assuieu the public that they hau leaineu fiom theii mistakes anu that safety was
a top piioiity. All of this was just five yeais piioi to the BP oil spill in the uulf of
Nexico Aftei the spill, much of BP's iesponses weie uefensive. In theii initial piess
ieleases, BP focuseu on how they weie the assisting paity iathei than the
9
iesponsible paity. The fiist piess ielease issueu on Apiil 21, 2u1u was a
confiimation anu copy of Tiansocean's ielease announcing the explosion (Biitish
Petioleum, p.1). BP's seconu piess ielease, also issueu on Apiil 21, 2u1u stateu that
BP touay offeieu its full suppoit to uiilling contiactoi Tiansocean Ltu.
Anu its employees aftei fiie causeu Tiansocean's submeisible uiilling
iig Beepwatei Boiizon to be evacuateu oveinight, saying it stoou
ieauy to assist in any way in iesponuing to the inciuent. (Biitish
Petioleum, paia. 1)
It was not until Apiil 24, 2u1u that BP even acknowleugeu publicly the loss of life
that occuiieu as a iesult of the spill (Biitish Petioleum, p. 1). BP iemaineu steaufast
that they weie in a suppoiting iole, anu not iesponsible foi the uisastei. In Bob
Buuley's Nay 6, 2u1u speech at the Chief Executives' Club of Boston (which was the
fiist speech aftei the spill), the event was blameu on a faulty blowout pieventei
(Buuley, 2u1u). As moie infoimation was ieleaseu, Ballibuiton's faulty cement job
woulu come into play as anothei guilty paity BP sought to blame. Eventually, it
came into question whethei oi not BP was awaie of the fact that the cement job on
the Beepwatei Boiizon well was faulty; with eviuence pointing to the fact that they
weie awaie. This was highly ciiticizeu by many people anu gioups. BP spent $Su
million uefenuing its iecoiu to the public. Nany felt that this was $Su million that
shoulu have been spent to clean up the spill anu aiu those who hau been affecteu.
Nuialiuhaian et al. (2u11) quote Chiis uiluez of the global public ielations fiim Bill
anu Knowlton as saying, "until the leak is stoppeu, no amount of auveitising anu PR
will help" (p. 227).
1u
4:&(01 7/ !"# %&"## 7, 1>+ A+(#1> 7/ C)#/ '7(*1 '"1"B+,*
Apiil 2u, 2u1u, will be ueeply etcheu into the minus of citizens of the uulf
Coast. The spill's impact on the enviionment anu economy was alieauy pioving to
be uisastious. Linua NcCauley (2u1u) states that citizens of the uulf Coast will be
affecteu by the oil spill with shoit anu long teim psychological uamage. In auuition
to inciuences of suiciue as a iesult of stiess causeu by the spill, iesponueis anu
iesiuents of the uulf aie at iisk foi ueveloping anxiety uisoiueis anu post-tiaumatic
stiess uisoiuei. Similai issues weie seen in the wake of the Exxon valuez oil spill.
NcCauley also cites eviuence that long-teim iesponueis have shown signs of
iiiepaiable psychological uamage (pp. S4 - S6). In the .,+93= K++6#= L+@* (2u11)
the psychological effects aie highlighteu again in moie uetail. This aiticle cites
stuuies anu statistics uemonstiating that a similai level of stiess was piesent in
inuiviuuals in uulf Coast states that weie uiiectly influenceu by the oil spill anu
those in states, such as Floiiua, that weie not as uiiectly impacteu. Accoiuing to
.,+93= K++6#= L+@* (2u11), a 0niveisity of Naiylanu anu 0niveisity of Floiiua
stuuy showeu that the oil spill was a huge public mental health ciisis anu suggests
that some iecoveiy effoits neeu to auuiess these issues (p. 1).
M@7 M>+79+1"0(# '9"*"* <7-+#*
Two mouels of ciisis iesponse may help to bettei unueistanu both BP's
iesponse stiategy anu the public's iesponse to that stiategy. Coombs' (2uu4)
Situational Ciisis Communication Theoiy examines the impact of piioi events anu
attiibutions baseu on those events on ciisis ieaction. Benoit's (1997) image
11
iestoiation appioach looks at the actual messages communicateu by an
oiganization aftei the ciisis has occuiieu.
N119"D)1"7, M>+79= (,- %"1)(1"7, '9"*"* '7:
,"0(1"7, M>+79= Situational Ciisis Communication Theoiy (SCCT) posits that an inuiviuual's
knowleuge of past ciisis events an oiganization is chaigeu with being iesponsible
foi is necessaiy to take into consiueiation when auuiessing futuie ciisis
communication planning (Coombs, 2uu4, p. 26S). Accoiuing to Coombs', as
"ieputational thieat incieases, the ciisis managei shoulu use iesponse stiategies
that uemonstiate the acceptance of iesponsibility foi the ciisis anu that auuiess
victim conceins" (p. 266). BP, uue to the natuie of theii inuustiy anu the Texas City
iefineiy uisastei, was in a position of high ieputational thieat as they went into
ciisis iesponse situation uuiing the 2u1u oil spill. The high ieputational thieat anu
past iecoiu of safety issues leauing to seiious uisasteis means that BP woulu likely
be helu iesponsible in some uegiee foi the situation, no mattei how much they
coulu have influenceu oi pieventeu the situation.
Attiibution Theoiy is connecteu to SCCT because it holus that "people will
make juugments about the causes of events, especially unexpecteu events with
negative outcomes" (Weinei, 198S, p. 28u) anu that theie aie ceitain attiibutions, oi
peiceptions, that will influence how the event is juugeu. In paiticulai theie aie thiee
uimensions that aie useu to foim attiibutions, as iepiesenteu in Table 2-1.
12
Table 2-1
Attiibution Theoiy
#$$%&'($&)* +,-&*&$&)*
Stability If the cause happens fiequently (stable)
oi infiequently (unstable)
Exteinal Contiol Whethei the cause of the event was
contiollable oi uncontiollable by
anothei peison
Peisonal Contiol ¡ locus The extent to which a peison can contiol
his oi hei own fate in the situation
Each attiibution impacteu how inuiviuuals juugeu BP befoie, uuiing, anu aftei the
oil spill. Stability was uebateu because of the uisastei that claimeu seveial lives just
S yeais pievious to the spill. Exteinal contiol was a majoi factoi uuiing the post-
uisastei sciamble to finu facts about what happeneu anu whethei oi not it coulu
have been pieventeu. The uistinctiveness of the situation was calleu to question as
BP's tiack iecoiu was examineu. Accoiuing to Coombs (2uu7), a ciisis is not
uistinctive when "an oiganization shows little consiueiation foi stakeholueis acioss
a numbei of uomains, not just in ciisis" (p.1S7). As moie infoimation became cleai
about the uetails of what happeneu anu BP was woiking to iesponu to the public
ielations ciisis, manageis shoulu have lookeu to Situational Ciisis Communication
Theoiy as a guiue to finuing the best possible way to iesponu, given theii ieputation
anu the situation.
'9"*"* '7:
,"0(1"7, I+*&7,*+ <7-+#* Theie aie two mouels that attempt to explain ciisis communication
stiategies anu iesponses. Benoit (1997) takes a moie ihetoiical look at the ciisis
1S
communication messages by focusing on the stiategic iationale behinu the message,
while Biaufoiu anu uaiiet (199S) look at a two-factoi appioach that focuses on a
iesponse oi no iesponse uichotomy.
4
3+ 9+*179(1"7, 1>+79= BP was left with the uaunting task of iestoiing its image to tiy anu get back
in the public favoi. This is a veiy uifficult public ielations task, especially when
uealing with a uisastei as laige anu uevastating as the oil spill. Benoit's (1997)
"Image Restoiation Theoiy" pioviues the theoietical ieasoning behinu why ceitain
stiategies aie best employeu in paiticulai situations. The theoiy is baseu on basic
concepts that piopose that the accuseu paity is iesponsible anu that the accusation
is consiueieu offensive (Benoit, 1997, p. 177). Table 2-2 piesents the 14 stiategies
that have emeigeu fiom Benoit's theoiy iange fiom simple uenial to moitification
(Bailow, 2u11, p! 81).
0sing Benoit's theoiy as a basis foi his stuuy, Bailow says that "BP was
focuseu on the twin tasks of uesciibing coiiective action anu uesciibing attempts to
compensate the victims of the inciuent" (p. 82). Fuithei, he states that BP nevei
boie any of the iesponsibility foi the actual explosion anu iesulting oil spill, iathei
they simply accepteu iesponsibility foi uealing with its afteimath. This is a point
that causeu significant tension (p. 82).
14
Table 2-2
Benoit's (1997) Image Restoiation Stiategies
.$%/$,01 23/456,
Simple uenial We uiu not uo it
Shifting blame Someone else uiu it
Piovocation We uiu it, but weie piovokeu
Befeasibility Lack of infoimation oi contiol
Acciuents The inciuent was an acciuent
uoou intention The eiioi was the iesult of goou intentions
Bolsteiing 0ui goou chaiacteiistics outweigh any eiioi
Ninimization The pioblem is not that bau
Biffeientiation The inciuent is uiffeient fiom some othei act
Tianscenuence
The act shoulu be unueistoou in a uiffeient
context
Attack accusei The peison blaming us is the one at fault
Compensation The victim will be compensateu
Coiiective Action We will fix the pioblem
Noitification
We aumit iesponsibility oi ask foi
foigiveness
1S
'79&79(1+ 07:
,"0(1"O+ 9+*&7,*+ :7-+# Ciisis communication is a uelicate task to hanule. Theie must be a ciisis
communication plan in place piioi to a ciisis occuiiing in oiuei to piopeily auuiess
anu combat the situation. Biaufoiu anu uaiiet's (199S) coipoiate communicative
iesponse mouel "focuses on the potential effectiveness of coipoiate communicative
iesponses to accusations of unethical oiganizational behavioi" (p. 876). Theii
mouel is compiiseu of two paths: the no iesponse path anu the coipoiate iesponse
path. The no iesponse path finus that uuiing times of ciisis, oiganizations in
question may offei no iesponse foi the allegations being biought. Biaufoiu anu
uaiiet aigue that the no iesponse path geneially leaus to obseiveis assuming guilt);
alteinatively, the coipoiate iesponse path holus that, by offeiing some iesponse, an
oiganization can influence the obseivei's peiception of the situation (p. 878). 0nuei
the coipoiate iesponse path falls foui iesponses that may be offeieu in times of
ciisis. Table 2-S piesents these iesponses anu the optimal situations foi theii use.
16
Table 2-S
Biaufoiu anu uaiiet's (199S) Response Nouel
7,85)*8, +,-&*&$&)* .&$(/$&)*
Benials Benials aie statements that
ueny the occuiience oi
existence of the
questionable event, oi ueny
that the accuseu
oiganization is the cause of
the event.
When an actoi can pioviue
eviuence that he¡she uiu not
commit an allegeuly unethical
action (commission situation),
this actoi shoulu use a uenial
communicative iesponse.
Excuses Excuses aie statements that
aigue that the accuseu
oiganization shoulu not be
helu iesponsible foi the
occuiience anu¡oi impact
of the questionable event
because ceitain factois
limiteu the oiganization's
contiol ovei the occuiience
anu¡oi impact of this event.
When an accuseu actoi can
pioviue eviuence that he¡she
uiu not have contiol ovei the
occuiience anu¡oi impact of
an allegeuly unethical action
(contiol situation), this actoi
shoulu use an excuse
communicative iesponse
}ustifications }ustifications aie statements
that aigue that, even though
the accuseu oiganizations is
iesponsible foi the
questionable event, the
stanuaius being useu by the
accuseis to evaluate the
impact of this questionable
event aie inappiopiiate.
When an accuseu actoi can
pioviue eviuence that
inappiopiiate stanuaius aie
being useu to evaluate this
allegeuly unethical action
(stanuaius situation), this
action shoulu use a
justification.
Concessions Concessions aie statements
that agiee that the
questionable event uiu
occui, that the accuseu
oiganization causeu this
event, that they evaluative
stanuaius being useu by the
accuseis aie appiopiiate.
When an accuseu actoi
concluues that the allegations
iaiseu aie valiu (agieement
situation), the actoi shoulu
use a concession.
Bue to the scope anu natuie of the BP uisastei, issuing a iesponse of some natuie
was necessaiy. Bowevei, in the beginning of theii communicative iesponse, BP uiu
17
what they coulu to tiy to shift the blame to othei playeis in the uisastei. This mouel
uemonstiates that the iesponse BP chose uiu not coiiesponu with the situation
cieateu by theii unethical behavioi. This likely playeu a iole in the level of public
acceptance of the communicative anu public ielations measuies BP took in the wake
of the uisastei.
N '>(,3",3 <+-"( .,O"97,:+,1
Two factois meuiate any analysis of BP's ciisis iesponse stiategy anu
communications. The fiist ueals with a changing communication meuium÷the iise
of a populist-appioacheu social meuia. The seconu ueals with a ieneweu awaieness
of the fiagility of the woilu's enviionment. Both help to bettei unueistanu BP's
stiategy anu its failuies at communicating.
%70"(# :+-"( 9+O7#)1"7, (,- 09"*"* 07:
,"0(1"7, The BP oil spill happeneu uuiing a time when two veiy impoitant
ievolutions weie happening. Fiist, social meuia was taking ovei anu changing the
way communications, public ielations, anu oiganizations functioneu. Wateis,
Tinuall, anu Noiton (2u1u), foi instance, examineu how public ielations has
iesponueu to anu evolveu with the incieaseu use of social meuia. They founu that
moie iepoiteis aie now open to getting news anu pitches thiough alteinative
souices. The stuuy goes so fai as to cite }ulia Boou, then euitoi at 8I K++6, as saying
that the "meuia pitch is ueau" (p. 242).
Reaching the meuia anu public is changing because of social meuia, foicing
iepoiteis, publicists, anu all communicatois to auapt. The meuia is no longei solely
18
getting stoiies thiough tiauitional pitches; iathei they aie looking at alteinative
methous, incluuing social meuia, to get stoiies. This is impoitant when stuuying the
BP oil spill because it was one of the fiist expeiiences with ciisis anu social meuia.
Buiing the Exxon valuez oil spill, public ielations piactitioneis uiu not have to
woiiy about Twittei anu Facebook anu Blogs. The BP oil spill bioke the molu anu
public ielations piofessionals hau to seiiously consiuei the iepeicussions of social
meuia anu how they woulu utilize it to combat the situation. It was also the fiist oil
spill wheieby the public hau access to social meuia. Repoiteis weie able to ieach
out the public on social meuia to get feeuback anu infoimation.
Social meuia, as it is available at peoples' fingeitips, has become influential in
how an inuiviuual foims opinions. Twittei, in paiticulai, has cieateu a sense of
community wheie inuiviuuals use 14u chaiacteis to post links, expiess opinions,
anu shaie infoimation with those who "follow" them anu aie pait of theii
community. The community peimits the shaiing of opinions anu values between
useis (Auei, 2u11, p. 711). Fuitheiing this, stuuies conuucteu just aftei the 2uu8
Piesiuential election in the 0niteu States showeu that among social meuia useis
ageu 18 - 29, Su% of the population polleu by Pew saiu they useu social meuia to
uiscovei, leain, anu influence theii voting uecision anu opinions on issues. Foi useis
age Su anu ovei, the peicentage was S2% (Auei, 2u11, pp. 711-712).
The piopensity of people to use social meuia as a place to gathei news uuiing
ciisis can be uemonstiateu by compaiing the iesponse of a pie-social meuia ciisis to
a cuiient ciisis. Buiing the teiioiist attacks of Septembei 11, 2uu1, officials hau to
spenu millions of uollais anu time to synchionize communication systems. In
19
compaiison, uuiing the Baiti eaithquake of 2u1u, Twittei was useu as a meaningful
channel of emeigency communication (Auei, 2u11 ,p.72S).
Social meuia has also incieaseu as souices foi news because of the gieatei
amount of fieeuom it allows foi. Wheie ceitain types of images oi stoiies aie
iestiicteu, social meuia has bioken the baiiiei anu alloweu foi all news anu stoiies
to be given aiitime (Cheong anu Lee, 2u1u, p. 47). This was uemonstiateu most
iecently uuiing the so-calleu "Twittei Revolution" anu othei social meuia leau
ievolts (Noiozov, 2u1u, pp. 24 - 2S).
Twittei has become a poweiful woiu of mouth tool to shaie sentiments anu
expeiiences, both positive anu negative, about bianus anu companies. }ansen,
Zhang, Sobel, anu Chowuuiy (2uu9) foi example founu that 19% of micioblogs
containeu a mention of a bianu oi company, anu of that peicentage, Su% weie
positive anu SS% weie negative (p. 2169). Twittei, as a micioblog, is accessible at
all times foi a usei to quickly post an opinion as it comes to top of minu. It enables
useis to immeuiately wiite how they feel oi what they think uuiing expeiiences.
Theii stuuy also founu that theie is a 0-shapeu ielationship between an inuiviuual's
expeiience anu theii likelihoou to expiess it (}ansen et al., 2uu9, p. 2178).
Twittei is aichival, which incieases its ielevance anu impoitance in
communication. Although theie aie limits to this on the actual Twittei website,
theie aie numeious ieliable aichive websites wheie you can access the past
(Fischei, 2u1u, p. 1S). 0nlike a face-to-face conveisation, wheie uetails can be lost
ovei time, Tweets can be founu anu tiaceu back anu founu in the context that they
weie wiitten. They have become pait of histoiy in how they have been useu in ciisis
2u
anu leau to ievolution. Social netwoiks, incluuing Twittei, have uemonstiateu theii
impoitance by allowing foi the continuation of piogiess anu thiough pioviuing an
"instantly accessible anu ongoing inventoiy of infoimation that can be useu to solve
pioblems anu even potentially to pioviue emotional suppoit" (NcSwite, 2uu9, p.
92).
M>+ .,O"97,:+,1(# I+O7#)1"7,
The seconu ievolution taking place uuiing the uulf oil ciisis was a shift of
focus towaius enviionmentalism. Caie foi the enviionment anu instiuction to uo so
can be tiaceu back to ancient times. It is in the G"D#+ anu P)9H(,. Theouoie Roosevelt
leu an enviionmental movement uuiing his piesiuency between 19u1 anu 19u9.
Although its histoiy uates back so fai, many people uiu not take into consiueiation
how theii actions effect the enviionment until moie iecently. Enviionmentalism has
become an impoitant pait of business anu especially maiketing uue to the fact that
costumeis aie now looking foi enviionmental anu social iesponsibility when
making puichases. Stuuies have founu that consumeis piefei "gieen" piouucts anu
woulu switch bianus anu spenu moie money in oiuei to puichase these piouucts
(Smith, 2u1u p. 2). Even piioi to the oil spill, the uamages causeu by oil anu gas
companies tuineu people off to these inuustiies anu theii images weie huiting.
Thus, the uegiee of ciisis facing BP was enhanceu uue to a ieneweu iesponse to
enviionmental uisasteis anu impacteu on not only what people communicateu
about the ciisis but also in new communication channels.
21
I+*+(90> P)+*1"7,*
Nost of the existing liteiatuie iegaiuing the 2u1u uulf 0il Spill examines the
causes, effects, anu outcomes of the uisastei. The complete outcome of the spill is
unknown anu may not be fully unueistoou foi sometime as inuustiy anu
enviionment both iequiie significant time to bounce back, anu fact finuing anu legal
battles iegaiuing the spill aie still ongoing. Existing social meuia liteiatuie laigely
focuses on pievalence of uiffeient platfoims anu how the powei of social meuia can
be hainesseu anu best useu to solve ceitain communication issues. uiven the extent
of existing ieseaich on social meuia anu liteiatuie iegaiuing ciisis communication,
it is possible to stuuy how ceitain stiategies weie employeu in BP's public ielations
iesponse following the spill. Baseu on the pieceuing uiscussion, the following
ieseaich questions aie poseu:
IPQR What was the uominant tone of Tweets uiiecteu towaius Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
IPSR What was the uominant focus of Tweets uiiecteu towaiu Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
IPTR What weie the uominant Image Restoiation Stiategy employeu by
Biitish Petioleum on Twittei in iesponse to the 2u1u uulf oil spill.
22
Chaptei S
Nethouology
In oiuei to stuuy both how Biitish Petioleum useu Image Restoiation
Stiategy (Benoit, 1997) in theii social meuia iesponse to the oil spill anu how the
public useu social meuia as a platfoim to iesponu anu ieach BP aftei the 2u1u uulf
oil spill, a content analysis was conuucteu. The analysis of Tweets geneiateu by
BP_Ameiica anu Tweets wiitten by the public to ÇBP_Ameiica weie all piouuceu
between Nay 2, 2u1u anu }anuaiy 1S, 2u12. The gieatest peicentages (89.S% anu
8S.S%) of content came fiom the fiist few months immeuiately following the oil
spill.
J,"1* 7/ N,(#=*"*
Seveial units of analysis weie analyzeu in oiuei to gathei all necessaiy
infoimation to answei the ieseaich questions. The units of analysis incluueu social
meuia type, message tone, genuei of senuei, uate of message, message focus, anu BP
image iestoiation stiategy (see Appenuix A foi Couebook).
%
&#",3 All Tweets useu in the samples foi content analysis weie selecteu at ianuom.
In oiuei to iuentify the vaiious image iestoiation stiategies useu by BP, a sample of
1u9 Tweets was selecteu at ianuom fiom the initial sample of Suu Tweets gatheieu
fiom an aichive seivice. A seconu sample on 82 Tweets uiiecteu at BP was
ianuomly selecteu fiom the initial gatheieu sample of 17u Tweets uiiecteu at BP
fiom an aichive seivice. In all, 191 Tweets compiiseu the sample useu to analyze
BP's iestoiation stiategies anu answei the thiee stateu ieseaich questions.
2S
'7-",3
Tweets piouuceu by the public uiiecteu towaius ÇBP_Ameiica weie coueu
foi the following vaiiables: uate (mm¡uu¡yy), genuei of the senuei, tone of the
message, anu focus of the message. Tweets piouuceu by BP_Ameiica weie coueu foi
the following vaiiables: uate (mm¡uu¡yy), image iestoiation¡not image iestoiation,
anu Benoit's 14 image iestoiation stiategies (See Table 2-2). Each vaiiable anu its
meaning was agieeu upon by the coueis (See Appenuix A ):
• +/$,: The yeai, month, anu uate that each inuiviuual tweet was
publisheu. Each Tweet in this stuuy hau a uate that shows when it
was publisheu.
• 9,*:,%: When possible, the genuei of the inuiviuual was iecoiueu as
"male" oi "female". At times, the genuei was uncleai anu iecoiueu as
"othei".
•
*,: When applicable, the Tweets weie analyzeu foi tone. Content that was wiitten in suppoit of BP oi its effoits to clean the uulf,
iectify the situation, oi take iesponsibility foi what they hau uone
weie classifieu as "positive". Content that conuemneu BP, its
measuies in the uulf, theii iesponse to the oil spill in geneial weie
categoiizeu as "negative". Also incluueu as negative weie any Tweets
that ciiticizeu BP's iesponse as ineffective, insufficient, oi impiopei.
Content that challengeu BP's honesty anu integiity in the cleanup oi
afteimath in any weie categoiizeu as negative. Some content
containeu both positive anu negative iemaiks towaius BP anu theii
24
iesponse to the oil spill anu as such was categoiizeu as "neutial".
Content that expiesseu no stiong feeling, offeieu suggestions oi
claiifications was also categoiizeu as neutial.
• <)=(8 )- $>, ?,88/0,: When possible, the coueis iuentifieu the main
focus of the Tweet as being in iegaius to inuustiy uamage oi
piogiess, health conceins, enviionmental uamage oi piogiess,
financial anu economic impact of the spill, iecoveiy effoits, oi othei.
• @4/0, 7,8$)%/$&)* .$%/$,01: When applicable, the coueis
categoiizeu the Tweets into one of the 14 uiffeient stiategies
incluueu in Benoit's (1997) Image Restoiation Stiategy theoiy (See
Table 2-2).
'7-",3 9+#"(D"#"1=
The ieseaichei anu anothei public ielations giauuate stuuent, aftei agieeing
on guiuelines anu instiuctions, coueu a test set of 1u7 Tweets uiiecteu to
ÇBP_Ameiica anu geneiateu by ÇBP_Ameiica anu ieacheu an acceptable level (.84)
of ieliability ueteimineu by using Bolsti's (1969) ieliability foimula.
%1(1"*1"0(# N,(#=*+*
Aftei couing ieliability was establisheu, the sample sets weie coueu anu all
uata was enteieu into SPSS foi uesciiptive anu infeiential statistical analyses.
Specifically, uata weie uisplayeu by fiequency anu peicent anu testeu foi
uiffeiences via Chi-squaie tests.
2S
Chaptei 4
Results
The puipose of the specific ieseaich questions that weie put foiwaiu in this
stuuy was to specifically inuentify how iestoiation stiategy was useu on Twittei,
how inuiviuuals useu Twittei as a platfoim to voice theii opinions in the wake of the
spill, anu finally to ueteimine the piimaiy concein voiceu by of Twittei useis aftei
the spill. Answeiing these questions pioviues bettei insight into how social meuia
can be best employeu in ciisis anu why BP's ciisis communication plan uiu not woik
out that well.
The stuuy conuucteu a content analysis of 191 Tweets geneiateu by oi
uiiecteu at Biitish Petioleum aftei the Apiil 21, 2u1u explosion on the Beepwatei
Boiizon uiilling iig anu the iesulting oil spill. All uata collecteu fiom the sample set
of Tweets weie analyzeu using Statistical Package foi the Social Science (SPSS).
?+*09"&1"O+ %1(1"*1"0*
The uata collecteu to answei the ieseaich questions was gatheieu fiom the
Twittei aichive seivice Topsy. 17u Tweets ÇBP_Ameiica weie collecteu to gathei a
sample size of 82 Tweets foi analysis. Analyzeu content came fiom uates beginning
on Nay 2, 2u1u, anu enuing on Septembei 14, 2u11. Although the Tweets foi the
analyzeu sample weie collecteu at ianuom, the majoiity of content (8S.S%) was
geneiateu in 2u1u fiom the beginning of the oil spill until August 2u1u, just aftei the
well was cappeu. In the afoiementioneu sample the genuei of the senueis was
iepoiteu as SS.4% male, 17.1% female, anu 47.6% othei (see Table 4-1).
26
Table 4.1 uenuei of Tweet Senuei
Gender
Frequency Percent Valid Percent
Cumulative
Percent
Valid Other 39 47.6 47.6 47.6
Male 29 35.4 35.4 82.9
Female 14 17.1 17.1 100.0
Total 82 100.0 100.0
Bata collecteu to answei the ieseaich questions in iegaiu to Tweets
geneiateu by Biitish Petioleum weie also gatheieu fiom Twittei aichive seivice
Topsy. Thiee hunuieu twenty Tweets fiom BP_Ameiica weie collecteu anu a sample
of 1u9 Tweets was selecteu at ianuom foi analysis. Analyzeu content ianges in uate
fiom Nay 1u, 2u1u, to }anuaiy 1S, 2u12. Although the Tweets weie selecteu at
ianuom, in the same mannei as the fiist sample, the majoiity (89.S%) of the Tweets
in the sample aie fiom the time span fiom the beginning of the spill to shoitly aftei
the well was cappeu. Almost eighty peicent (79.8%) of the content in the sample
was iestoiation stiategy, while 2u.2% of the content was not iestoiation stiategy
(see Table 4-2). Tweets coueu as not iestoiation stiategy simply containeu some
basic infoimation oi uiu not peitain to the oil spill anu iecoveiy.
Table 4.2 Peicentage of Tweets Containing Restoiation Stiategy
Frequency Percent Valid Percent
Valid Restoration 87 79.8 79.8
Not Restoration 22 20.2 20.2
Total 109 100.0 100.0
27
M+*1 7/ I+*+(90> P)+*1"7,*
IP QR What was the uominant tone of Tweets uiiecteu towaius Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
In oiuei to ueteimine the uominant tone of the Tweets uiiecteu towaius BP,
each inuiviuual unit of analysis was coueu to ueteimine whethei the oveiall tone of
the message was positive, negative, oi neutial (X
2
=47.7S, uf=2, p=u). As founu in
Table 4-2, iesults showeu that of the 1u9 tweets analyzeu that 6S.4% (S2) weie
negative in tone. The uefinition of "negative" as useu in the analysis was any content
that conuemneu Biitish Petioleum, oi any content that ciiticizeu BP's iesponse as
ineffective, insufficient, oi impiopei. Foi example, negative Tweets incluueu one
fiom Nay 19, 2u1u when Çgeoffliving Tweeteu "It seems like ÇBP_Ameiica claim to
Sk baiiels a uay in the oil spill is the biggest PR lie of the yeai so fai. #BPheaiing".
Neutial-toneu Tweets accounteu foi SS.4% (29) of the analyzeu content.
"Neutial" was uefineu as content that containeu an equal amount of positive anu
negative, content that uiu not expiess stiong feeling, anu content that gave
suggestions, claiifications. Content coueu as neutial incluueu a Tweet fiom Nay S,
2u1u by Çzaibatsu that says "Let ÇBP_Ameiica know the #0ilSpill is impoitant to
you" anu anothei Tweet fiom }uly 1, 2u1u when PBS News Boui (ÇNewsBoui)
Tweeteu, " Watch ÇBP_Ameiica inteiview live on oui Çustieam". The final 1.2%
(1) of the content was positive in tone. In oiuei foi a Tweet to be "positive," it hau to
contain content that was in suppoit of BP anu theii effoits to clean the uulf oi in
28
piaise of the piogiess maue by BP in the uulf. The only iecoiueu positive Tweet was
a Tweet that stateu that BP hau biought the community togethei in the uulf foi a
laige baibeque event.
Table 4.S Recoiueu Tone of Tweets Biiecteu Towaius BP
Frequency Percent Valid Percent
Valid Negative 52 63.4 63.4
Neutral 29 35.4 35.4
Positive 1 1.2 1.2
Total 82 100.0 100.0
Chi-squaie value: 47.7S
IP SR What was the uominant focus of Tweets uiiecteu towaiu Biitish
Petioleum uuiing anu aftei the 2u1u uulf oil spill.
To ueteimine the uominant focus of Tweets uiiecteu towaiu Biitish
Petioleum in the uata set, content was coueu anu categoiizeu in one of six possible
categoiies, incluuing: inuustiy, health, enviionment, financial anu economic impact,
iecoveiy, anu othei. The puipose of this ieseaich question is to ueteimine what the
uominant conceins uuiing the oil spill weie among inuiviuuals anu gioups who
useu Twittei to communicate conceins anu issues uuiing the oil spill anu clean up
(see Table 4-4).
"0thei" accounteu foi the laigest peicentage, S9% (S2) of content. Nany of
the Tweets that fell into this categoiy containeu messages that expiesseu fiustiation
as whole towaius BP, iathei than conceining a specific focus. The enviionment was
the most focuseu on specific categoiy, with Su% (2S) of the tweets containing
content that conceineu what BP was uoing to fix the enviionment oi othei content
29
that was in iegaius to enviionmental impact uue to the spill. Tweets iegaiuing the
iecoveiy piocess compiiseu 2u.7% (17) of the analyzeu content. 4.9% (4) of Tweets
weie in iegaiu to the financial anu economic impact that the spill hau on those that
it hau affecteu. Inuustiy anu health both weie the focus of 2.4% (2) of the content
that was analyzeu to answei the above stateu ieseaich question.
Table 4.4 Recoiueu Focuses of Nessages
Frequency Percent Valid Percent
Valid Industry 2 2.4 2.4
Health 2 2.4 2.4
Environment 25 30.5 30.5
Financial / Economic impact 4 4.9 4.9
Recovery 17 20.7 20.7
Other 32 39.0 39.0
Total 82 100.0 100.0
Chi-squaie value: 61.S6, uf=S, p=u
To expanu on the fiist two ieseaich questions, cioss tabulation was
calculateu to ueteimine any iiiegulai oi extieme vaiiances between tone anu focus,
genuei anu tone, anu genuei anu focus (see Tables 4-6, 4-6, anu 4-7).
Su
Table 4.S Tone - Focus Cioss Tabulation
Focus
Total Industry Health Environment
Fin. and econ.
impact Recovery Other
Tone Negative 1 1 20 3 10 17 52
Neutral 1 1 5 1 6 15 29
Positive 0 0 0 0 1 0 1
Total 2 2 25 4 17 32 82
Chi-squaie value: 8.928
Table 4.6 uenuei - Tone Cioss Tabulation
Tone
Total Negative Neutral Positive
Gender Other 23 15 1 39
Male 21 8 0 29
Female 8 6 0 14
Total 52 29 1 82
Chi-squaie value: 2.SuS
Table 4.7 uenuei - Focus Cioss Tabulation
Focus
Total Industry Health Environ.
Fin. and econ.
impact Recovery Other
Gender Other 1 1 13 2 10 12 39
Male 1 1 9 1 2 15 29
Female 0 0 3 1 5 5 14
Total 2 2 25 4 17 32 82
Chi-squaie value: 8.2S
The cioss tabulation tables uemonstiate that theie is vaiiance between tones
anu focus (X
2
=8.9S, uf=1u, p=.SSS). The aieas of focus that weie most pievalent in
the sample also show the highest concentiation of negativity. Tweets piouuceu by
S1
females appeai to be moie evenly split between negative anu neutial content than
Tweets piouuceu by males (X
2=
2.S1, uf=4, p=.64)
. The genuei - focus cioss
tabulation (X
2=
8.2S, uf=1u, p=.6u4) shows some unbalance in the uistiibution of
message focus between the genueis. Women show a highei concentiation of Tweets
focuseu on iecoveiy while men hau a highei focus on Tweets focuseu on the
enviionment. The highei peicentage of males in the sample, in auuition to the high
fiequency of Tweets that weie unable to be categoiizeu into one of the five aieas of
focus likely may have skeweu the iesults in the paiticulai aiea.
IP TR What weie the uominant Image Restoiation Stiategy employeu by
Biitish Petioleum on Twittei in iesponse to the 2u1u uulf oil spill.
0ne hunuieu nine Tweets geneiateu by Biitish Petioleum weie analyzeu to
ueteimine the uominant Image Restoiation Stiategy (Benoit, 1997) employeu by BP
in Twittei content aftei the spill anu uuiing the iecoveiy fiom the uisastei
(X
2
=169.26, uf=9, p=u). As seen in Table 4-7, almost 8u% (87) of the Tweets weie
iestoiation stiategies anu able to be categoiizeu in nine of Benoit's possible 14
stiategies. 2u% (22) Tweets weie not image iestoiation, meaning that they weie
just some foim of simple infoimation anu unable to be categoiizeu in one of the
possible 14 stiategies. Foi example, on }uly 7, 2u1u ÇBP_Ameiica Tweeteu an
announcement encouiaging followeis to "stay tuneu foi live tweeting of the latest
upuates." This Tweet is just a way of infoiming followeis that moie infoimation will
be ieleaseu in the neai futuie, it is not a iestoiation tactic. Anothei example of
S2
content that was coueu as not iestoiation is fiom }uly 14, 2u1u when ÇBP_Ameiica
Tweeteu about a biiefing being moveu up by an houi.
Foity-two peicent (46) of the Tweets weie coueu as coiiective action, which
was the most common stiategy employeu by BP. Coiiective action Tweets focuseu
on technical measuies that weie being taken by BP to stop the spill anu clean the oil.
0n }uly 1S, 2u1u ÇBP_Ameiica Tweeteu "At 12:Supm we staiteu to close the choke
- we uiu it half a tuin at a time to slowly close the well in." This Tweet was coueu as
coiiective action anu is fiom the uay that the well was finally sealeu. Anothei
example of coiiective action was Tweeteu on Nay 1u, 2u1u when ÇBP_Ameiica
saiu "We aie also in the piocess of gaining access at the bottom of the B0P. The next
step is the junk shot anu top kill to shut off the flow."
Bolsteiing was the seconu most common stiategy anu encompasseu almost
16% (17) of the coueu content. Bolsteiing was uefineu as when BP tiieu to
emphasize the positive things they weie uoing iathei than the negative events that
weie actually taking place. Foi example, BP often Tweeteu about seafoou safety anu
how veiy few tests on uulf sea foou weie coming back with positive iesults foi
contamination. This iecuiient theme uemonstiates bolsteiing because while all the
seafoou was supposeuly safe, the well was still not capeu anu oil was still flowing.
0f the iest, 9.2% (1u) weie classifieu as using compensation as iestoiation
stiategy; tianscenuence was employeu in 4.6% (S) of the analyzeu uata;
minimization as stiategy was useu in 2.8% (S) in the uata; shifting blame anu goou
SS
intention weie each obseiveu 1.8% (2) of the time; anu simple uenial anu attacking
the accusei weie each employeu in .9% (1) of the content .
Table 4.8 Image Restoiation Stiategy Fiequencies
Frequency Percent Valid Percent
Valid Not Restoration Strategy 22 20.2 20.2
Simple Denial 1 .9 .9
Shifting Blame 2 1.8 1.8
Good Intention 2 1.8 1.8
Bolstering 17 15.6 15.6
Minimization 3 2.8 2.8
Transcendence 5 4.6 4.6
Attack the Accuser 1 .9 .9
Compensation 10 9.2 9.2
Corrective Action 46 42.2 42.2
Total 109 100.0 100.0
Chi-squaie value: 169.26
S4
Chaptei S
Biscussion
Social meuia has become incieasingly impoitant in the past uecaue as moie
platfoims have become available foi useis to connect with othei useis anu geneiate
theii own content. It is now expecteu that an inuiviuual, oiganization, oi business be
piesent anu active to some uegiee on social meuia platfoims. Thiough cieating
connections between inuiviuuals anu oiganizations, social meuia has changeu the
business climate. An inuiviuual can now seek costumei seivice via social meuia
iathei than the olu fashioneu way of making a phone call. In that way anu some
otheis, social meuia has incieaseu the efficiency of oiganizations anu connecting
with the public. Bowevei, it has also alloweu foi inuiviuuals to have a platfoim to
spieau messages of uiscontent oi bau expeiiences to all of theii "followeis" anu
"fiienus" anu those Tweets, posts, anu viueos have uemonstiateu theii ability to
cause so much uamage by going viial oveinight. Anothei change social meuia is in
pait iesponsible foi is the call foi moie tianspaiency. Inuiviuuals now uemanu to
know how a company is opeiating anu how theii policies will affect them, anu social
meuia has cieateu a connection between inuiviuual anu coipoiations to facilitate
this exchange in infoimation.
In Apiil 2u1u, when oil began to poui into the uulf anu ieach the shoies in
numeious coastal communities, the effecteu inuiviuuals, inuustiy leaueis, anu
auvocates weie iightly infuiiateu with Biitish Petioleum. The tiuth about who is to
blame foi the spill anu the total uamage is still unknown anu evolving as moie
infoimation is ieleaseu anu stuuies aie conuucteu. Theie aie many lessons to be
SS
leaineu fiom the oil spill, anu one of those lessons is about the use of social meuia in
times of ciisis. Inuiviuuals can ieach an oiganization anu voice conceins,
fiustiations, anu suggestions in a public foium like nevei befoie. Similaily, Biitish
Petioleum was able to use social meuia as a platfoim to auuiess conceins anu
fiustiations while ieaching mass amounts of people with infoimation to help
contiol the ciisis, if uone piopeily.
This stuuy focuseu solely on the social meuia platfoim Twittei anu how
inuiviuuals useu it to connect with BP anu how BP, in tuin, useu it to ieach
inuiviuuals anu tiy to iecovei fiom the public ielations ciises that ensueu as a iesult
of the explosion anu spill. Twittei allows foi a iapiu way to engage because it's
accessible anu fiee of chaige to anyone who has access to the Inteinet anu is
equippeu with a computei oi othei uevice. It can be a fast anu efficient way to ielay
vital infoimation in times of ciisis. Twittei allows foi two-way symmetiical
communication, a two-way conveisation wheie connections aie foigeu anu
feeuback can be given anu ieceiveu (uiunig, 199u, p. 21).
At the time of the 2u1u oil spill, many companies anu oiganizations weie
opeiating at some level on social meuia. Whethei those company leaueis weie
awaie of it oi not, infoimation about the company coulu likely be founu somewheie
on social meuia platfoims. BP, piioi to the oil spill, hau a ielatively inactive Twittei
account . That is, they uiu not use Twittei to actively engage with the public until
aftei the oil spill. In fact, aftei the spill, Twittei accounts such as ÇBPulobalPR
became populai anu gaineu moie followeis than the veiifieu BP account. It seems
S6
cleai that BP was ieactive iathei than pioactive in theii use of social meuia, which
just fuithei uemonstiates theii lack of ciisis communication plan piepaieuness.
Piioi to the ciisis, Biitish Petioleum was not in the best giaces in the eyes of
many uue to the natuie of the inuustiy in which they opeiate. Big oil uiu not neeu a
massive ciisis to be hateu. 0il plays a laige iole in uomestic politics anu foieign
policy. At the time of the spill oil piices hau been soaiing anu many families anu
inuiviuuals in the 0niteu States weie stiuggling financially as a iesult.
Enviionmentalists anu the incieasing numbei of inuiviuuals who, at the time, weie
beginning to view oil as "uiity" uue to the impact of off-shoie uiilling weie calling
foi big oil anu the goveinment to uevelop anu funu cleanei eneigy even befoie the
spill. All of these factois, combineu with the 2uuS Texas City, Texas, explosion, left
BP in a situation of high ieputational thieat going into the ciisis causeu by the 2u1u
explosion anu spill (Coombs, 2uu4).
Consiueiing theii high ieputational thieat at the time of the ciisis, BP shoulu
have auopteu "stiategies that inuicate a gieatei acceptance of iesponsibility foi the
ciisis anu simultaneously uemonstiate concein foi victims" (Coombs, 2uu4, p. 269),
that is, employing social meuia channels to get its siue of the stoiy out anu cieate
both naiiative anu uialogue with its publics. BP showeu lack of piepaieuness anu
woulu have be able to cieate a uiffeient outcome foi themselves hau they taken the
time to ieseaich theii best options anu piepaie. Bau they iealizeu anu accepteu
theii ieputation anu position befoie the spill anu leain fiom the past, they may have
iecognizeu the impoitance anu potential benefits of taking a moie sinceie path.
Focusing on people, the enviionment, anu uamage to the inuustiy iathei than
S7
science anu extiemely technical uetails of the spill woulu have maue them bettei off.
Auuitionally, moie sinceiity fiom CE0 Tony Baywaiu anu an actual apology anu
aumittance of eiioi may have appeaseu fai moie people than the ieactive stiategy
that focuseu on technicalities aftei the spill.
The finuings of this stuuy aie consistent with pievious liteiatuie anu stuuies
about best piactices in social meuia anu best piactices in ciisis communications. BP
uiu not use piopei ciisis communication, anu that was uemonstiateu in theii
seemingly caieless iesponse. They employeu social meuia as a 9+(01"O+ methou anu
uiu not effectively engage anu connect with those using social meuia to ieach out
anu engage with them. They focuseu on veiy technical coiiective actions that they
weie taking to cap the well, stop the flow of oil, anu clean up the mess. 0ften times,
the language that was useu in Tweets was fai too technical foi the aveiage peison
who is not in the inuustiy to unueistanu. Foi example, a Tweet fiom }uly 1S, 2u1u
says "At 1u:Suam we closeu the kill line. At that point the only flow was going out
the choke line." Technical language like that can be uncleai to an aveiage peison
unless they hau knowleuge of what a kill like anu choke line aie. At times uuiing the
spill, BP even useu social meuia to makes jokes about the situation, incluuing one
fiom Nay 27, 2u1u about clean up iesuming as soon as a XXL wet suit was founu foi
a laigei inuiviuual iesponsible foi clean up — a stiategy that cleaily shoulu have
been avoiueu if they weie conceineu about theii image in any way.
Also in accoiuance with pievious liteiatuie was this stuuy's finuings about
the way inuiviuuals useu social meuia to ieach anu connect with BP. The gieatest
peicentage (6S.4%) of Tweets uiiecteu at ÇBP_Ameiica in this stuuy was negative.
S8
Consiueiing the giavity of the situation anu that most people have been founu to
expiess a stiong feeling in eithei uiiection when auuiessing a company oi piouuct
on social meuia, the laige negative social meuia iesponse was to be expecteu. The
laige peicentage of negativity can also be attiibuteu to be ieactive to BP's use of
Twittei uuiing the ciisis. Theii lack of willingness to engage with social meuia useis,
in auuition to theii iesponse stiategies, woikeu against them as eviuenceu by the
finuings in this stuuy.
%):
9= Social meuia has changeu the way inuiviuuals communicate. Two-way
communication thiough methous othei than meeting face-to-face oi thiough
telephones is now the noim. Communication platfoims like Twittei anu othei social
meuia have biiugeu a gap between inuiviuuals anu companies. The neeu foi
tianspaiency, honesty, anu connection is gieatei than evei because of how
accessible eveiyone has become. It is no longei necessaiy to wiite a lettei oi place a
phone call to voice youi opinion about a company oi situation. Piess ieleases, piess
confeiences, anu tiauitional methous of public ielations aie now just pait of the
piactice of builuing anu maintaining ielationships with the public because of social
meuia.
Buiing times of ciisis, social meuia has become a vital outlet to gathei anu
senu infoimation. It has become expecteu foi an oiganization, inuiviuual, oi
company to use social meuia to auuiess conceins anu uisseminate infoimation
uuiing times of ciisis whethei they aie emeigency ciises, coipoiate ciises, technical
malfunctions, oi otheiwise. BP's slow ieaction on social meuia anu unwillingness to
S9
engage with conceineu citizens looking to engage on Twittei anu othei social meuia
was just fuel to the fiie uuiing the oil spill anu its clean up.
Although the well has been cappeu in the uulf anu billions of uollais have
been spent on the clean up anu on compensation measuies foi citizens in the uulf,
BP still faces haish ciiticism foi theii hanuling of the oil spill. As time passes anu
moie infoimation comes out about BP piactices befoie anu uuiing the spill, in
auuition to the amount of uamage uone to the enviionment, inuustiy, anu health of
uulf citizens, BP has to continue to auuiess the ciisis anu piactice ciisis
communication. Theii use of social meuia has evolveu fiom being neaily non-
existent to being active on many platfoims anu also uemonstiating slightly moie
willingness to engage.
4:&#"0(1"7,*
Theie aie seveial implications that iesult fiom this stuuy. The fiist is that the
stuuy uemonstiates the neeu foi companies anu oiganizations to have a ciisis
communication plan in place that caiefully selects the stiategies that they will use in
accoiuance with best piactices. Some ieseaich woulu have given eviuence to
suppoit that BP shoulu have been less technical anu moie apologetic anu sinceie in
the wake of the uisastei. Some foiethought about using social meuia in the ciisis
communication plan woulu have alloweu them to be moie pioactive iathei than
ieactive anu moie engaging iathei than using the platfoim to uistiibute technical
infoimation. Ciisis communication has uemonstiateu in the past how it can be useu
to tiansfoim situations anu allow embattleu oiganizations to oveicome anu move
on aftei the ciisis enus. BP uiu not have to look to fai to see what woiks anu what
4u
uoes not woik in ciisis communication. Although the situation was unique in many
ways, the inuiviuuals iesponsible foi the ciisis communication plan coulu have
lookeu to the past. Specifically, they coulu have avoiueu Exxon's mistakes.
Anothei implication fiom this stuuy foi public ielations piofessionals
incluues knowing the impoitance of knowing youi auuience anu theii conceins. Bau
BP known theii auuience anu gaugeu theii conceins, they coulu have uiiectly
auuiesseu them. Social meuia allows foi monitoiing what people aie saying about a
paiticulai company thiough channels such as blogs anu Tweets. Taking the time to
seaich social meuia anu engage those with conceins, ciiticisms, oi even suppoit
helps bettei iuentify conceins anu auuiess them. BP's Twittei account uiu not
miiioi the conceins of those tiying to connect with them. The social meuia in
geneial, anu Twittei in paiticulai, aie moie than platfoims foi enteitainment; they
have uemonstiateu theii seiious implications anu powei anu shoulu not have been
oveilookeu anu misuseu as they weie by BP. BP's negligent behavioi was calleu to
attention in a Tweet fiom ÇNackcolliei who saiu "Seiiously if ÇBP_Ameiica uoesn't
caie enough to monitoi Twittei oi engage heie, what makes us think they caie
about ÇBPulobalPR.".
This stuuy showeu how inuiviuuals useu social meuia in times of ciisis to
voice theii opinions. In this case, theie was no shoitage of Tweets haishly ciiticizing
eveiy move that BP was making. 0n Nay S, 2u1u ÇNatuialPapa Tweeteu "Bey
ÇBP_Ameiica - why aie citizens volunteeiing to clean youi fieakin' mess. Put those
piofits to goou use anu clean up aftei youiself." Fuithei a Tweet fiom }une 4, 2u1u
geneiateu oiiginally fiom ÇLuxuiyPRual but ie-Tweeteu numeious times stateu
41
"BP is spenuing $Su NILLI0N on a new au campaign. They shoulu spenu that on
cleaning up the ocean. RT if u agiee ÇBP_Ameiica." Stiategic planning anu piopei
employment of social meuia in the futuie can allow foi public ielations piactitioneis
to achieve bettei outcomes than the outcomes ieacheu by the BP public ielations
plan.
2":"1(1"7,* (,- ?"9+01"7, /79 U)91>+9 I+*+(90>
This stuuy focuseu only on one aspect of social meuia, Twittei, anu its use in
ciisis communications. The iesults of this stuuy uemonstiate how BP's use of Image
Restoiation stiategy was useu in this one singulai outlet, anu how the public
ieacteu to the situation on just that one outlet. While social meuia is veiy impoitant,
it was just one aspect of BP's ciisis iesponse anu it was only one way that the public
ieacteu aftei the Apiil 2u1u uulf 0il Spill. Twittei is a populai way to vent
fiustiations anu feelings, but the foimat of the content uoes make it veiy uifficult to
inteipiet feeling as some things like saicasm anu humoi aie not easily unueistoou
in ceitain contexts.
Fuithei ieseaich shoulu incluue othei social meuia outlets to ueteimine if
theie was any vaiiance between Image Restoiation Stiategies useu acioss the
platfoims anu also to ueteimine if social meuia useis tone vaiieu acioss the
uiffeient platfoims in the messages uiiecteu towaius BP. Noie ieseaich may also
incluue iuentifying any uiffeience in tones anu message ieception in uiffeient
geogiaphic aieas. This stuuy uoes not take into consiueiation wheie the message
senuei lives anu whethei oi not they weie uiiectly affecteu by the spill oi not.
42
Fuithei ieseaich in these paiticulai aieas coulu iuentify vaiiances anu allow foi
bettei, moie thoiough ciisis communication planning in the futuie.
4S
Refeiences
Auei, N.R. (2u11). The policy sciences of social meuia. M>+ 87#"0= %1)-"+* V7)9,(#5
TWXYZ[ 7u9 - 7S2.
Benoit, W.L. (1997). Image iepaii uiscouise anu ciisis communication. 8)D#"0
I+#(1"7,* I+O"+@5 ST\S[ 177 - 184.
Biaufoiu, }.L. & uaiiett, B.E. (199S). The effectiveness of coipoiate communicative
iesponses to accusations of unethical behavioi. V7)9,(# 7/ G)*",+** .1>"0*5 QY5
87S - 892.
Biitish Petioleum. (2u1u). BP confiims that Tiansocean Ltu issueu the following
statement touay (Piess Release). Retiieveu fiom http:¡¡bp.com.
Biitish Petioleum. (2u1u). BP offeis full suppoit to Tiansocean aftei uiilling iig fiie
(Piess Release). Retiieveu fiom http:¡¡www.bp.com.
Biitish Petioleum. (2u1u). BP offeis sympathy to the families of those lost in the 0S
oil iig fiie (Piess Release). Retiieveu fiom http:¡¡www.bp.com.
Cheong, N. & Lee, v.C.S. (2u11). A micioblogging-baseu appioach to teiioiism
infoimatics: Exploiation anu chionicling civilian sentiment anu iesponse to
teiioiism events via Twittei. 4,/79
1"7, %=*1+: U97,15 QT[ 4S - S9. Chess, C. (2u1u). BP's iisk communication uisastei. .)97&+(, V7)9,(# 7/ I"*6
I+3)#(1"7,5 T:211 - 217.
Coombs, T.W. (2uu7). Attiibution theoiy as a guiue foi post-ciisis communication
ieseaich. 8)D#"0 I+#(1"7,* I+O"+@5 TT[ 1SS - 1S9.
Coombs, T.W. (2uu4). Impact of past ciises on cuiient ciisis communication:
Insights fiom situational ciisis communication theoiy. V7)9,(# 7/ G)*",+**
'7:
,"0(1"7,5 YQ[ 26S - 289. Buuley, B. (2u1u, Nay). G7D ?)-#+= (1 1>+ 0>"+/ +E+0)1"O+*H 0#)D 7/ G7*17,\ Speech
piesenteu at the Chief Executives' Club of Boston, Boston, Na.
Euosomwan, S., Piakasan, S.K., Kouame, B., Watson, }., & Seymoui, T. M>+ V7)9,(# 7/
N&&#"+- <(,(3+:+,1 (,- .,19+&9+,+)9*>"&5 S]QQ5 Q^[ 79 - 91.
Fischei, E. & Reubei, R.A. (2u11). Social inteiaction via new social meuia: (Bow) can
inteiactions on Twittei affect effectual thinking anu behavioi. M>+ V7)9,(# 7/
G)*",+** F+,1)9",35 S^[ 1 - 18.
uiattan, L. & Reese, A. (2u11). Bepiession: Bioauei psychological impact of 2u1u BP
oil spill. .,+93= K++6#= L+@*5 <(90> Y5 S]QQ[ S8S - S86.
44
uiiggs, }.W. (2u11). BP gulf of mexico oil spill. .,+93= G(9 N**70"(1"7, .,+93= 2(@
V7)9,(#5 TS: 1 - 12.
uiunig, }.E. (199u). Theoiy anu piactice of inteiactive meuia ielations. 8)D#"0
I+#(1"7,* P)(91+9#=5 T_[ 18 - 2S.
Bailow, W.F., Biantley, B.C., & Bailow, R. (2u11). BP initial image iepaii
stiategiesaftei the Beepwatei Boiizon spill. 8)D#"0 I+#(1"7,* I+O"+@5 T`[ 8u -
8S.
Bolsti, 0.R. (1969). '7,1+,1 (,(#=*"* /79 1>+ *70"(# *0"+,0+ (,- >)
,"1"+*\ Reauing, NA.: Auuison - Wesley Publishing Company.
}ansen, B.}., Zhang, N., Sobel, K. & Chowuuiy, A. (2uu9). Twittei powei: Tweets as
electionic woiu of mouth. V7)9,(# 7/ 1>+ N:+9"0(, %70"+1= /79 4,/79
1"7, (,- M+0>,7#73=5 ^]XQQZ[ 2169 - 2188.
NcCauley, L. (2u1u). Enviionments anu health: Will the BP oil spill affect oui health.
N:+9"0(, V7)9,(# 7/ L)9*",35 QQ][ S4 - S6.
NcSwite, 0.C. (2uu9). The challenge of social netwoiks. N-:","*19(1"O+ M>+79= (,-
89(E"*5 TQXQZ[ 78 - 9S.
Noiozov, E. (2u1u). Cognitive suiplus: Cieativity anu geneiosity in a connecteu age.
G7*17, I+O"+@5 T_[ 24 - 26.
Nuialiuhaian, S., Billistone, K., & Shin, }. (2u11). The gulf coast oil spill: Extenuing
the theoiy of image iestoiation uiscouise to the iealm of social meuia anu
beyonu petioleum. 8)D#"0 I+#(1"7,* I+O"+@5 T`: 226 - 2S2.
Ritchie, L.A., uill, B.A., & Picou, }.S. (2u11). The BP uisastei anu an Exxon valuez
ieiun. N:+9"0(, %70"7#73"0(# N**70"(1"7,5 %)::+9 S]QQ[ Su - SS.
Smith, L.C., Smith, L.N., & Ashcioft, P.A. (2u1u). Big oil, big consequences, anu the
big unknown: Exploiing the legal, iegulatoiy, anu enviionmental impact of
the uulf oil spill. N#D(,= 2(@ I+O"+@5 `Y[ 1 - 24.
Stacks, B.W., & Nichaelson, B. (2u1u). N &9(01"1"7,+9H* 3)"-+ 17 &)D#"0 9+#(1"7,*
9+*+(90>5 :+(*)9+:+,15 (,- +O(#)(1"7,. New Yoik: Business Expeit Piess.
Wateis, R.B., Tinuall, N.T.}., & Noiton, T.S. (2u1u). Neuia catching anu the jouinalist-
public ielations piactitionei ielationship: Bow social meuia aie changing the
piactice of meuia ielations. V7)9,(# 7/ 8)D#"0 I+#(1"7,* I+*+(90>5 SSXTZ: 241 -
261.
4S
Weinei, B. (198S). A)
, :71"O(1"7,\ New Yoik: Spiingei 46
APPENBIX A
;>,8&8 A):&*0
B*&$8 )- #*/618&8C
The stuuy examineu social meuia messages by Biitish Petioleum (BP) anu otheis.
Seveial units of analysis will be examineu in iesponse to the social meuia content
geneiateu by BP Ameiica as a iesponse to the oil spill.
.)=&/6 ?,:&/ ;15,C
D! Comments geneiateu by the public towaiu ÇBP_Ameiica on Twittei
E! Twittei content geneiateu by ÇBP_Ameiica
7,8$)%/$&)* .$%/$,01C
D! Restoiation stiategy
E! Not iestoiation stiategy
@4/0, 7,8$)%/$&)* .$%/$,01 FG,*)&$H DIIJKC
D! .&456, +,*&/6 - uefineu as when the accuseu paity iesponus to any
accusations by iefuting claims of iesponsibility
E! .>&-$&*0 G6/4, - uefineu as when the accuseu paity iesponus to
accusations by placing culpability on anothei paity that may oi may not
have been involveu
L! M%)N)=/$&)* - uefineu as when the accuseu paity aumits they aie to
blame, but that they weie piovokeu into committing the haimful action
O! +,-,/8&'&6&$1 - uefineu as when the accuseu paity iesponus to
accusations by saying that they lackeu contiol oi infoimation neeueu to
pievent the actions they aie accuseu of
47
P! #==&:,*$8 - uefineu as when the accuseu paity iesponus to accusations
by stating that the event was unintentional
Q! 9)): @*$,*$&)* - uefineu as when the accuseu paity iesponus to
accusations by explaining that the event was the iesult something
planneu with goou puipose that went wiong inauveitently
J! G)68$,%&*0 - uefineu as when the accuseu paity highlights positive
chaiacteiistics they holu which they believe outweigh the bau in the
situation
R! ?&*&4&S/$&)* - uefineu as when the accuseu paity uownplays the
seiiousness oi giavity of the situation
I! +&--,%,*$&/$&)* - uefineu as when the accuseu paity iesponus to the
accusations by explaining that the event is uiffeient fiom othei acts anu
cannot be compaieu to oi unueistoou thiough othei events
DT! ;%/*8=,*:,*=, - uefineu as when the accuseu paity iesponus to
accusations by explaining that the event shoulu be unueistoou in a
uiffeient context than it is cuiiently being unueistoou
DD! #$$/=U #==(8,% - uefineu as when the accuseu paity tuins the blame
aiounu anu puts it on the paity that is accusing them of causing the
pioblem
DE! A)45,*8/$&)* - uefineu as when the accuseu paity iesponus to the
accusations by stating that those impacteu will be able to iecovei
uamages, but uoes not necessaiily mean that the accuseu is takings
iesponsibility foi causing the uamage
48
DL! A)%%,=$&N, #=$&)* - uefineu as when the accuseu paity takes
iesponsibility foi unueitaking necessaiy measuies to fix the pioblem,
but uoes not necessaiily take iesponsibility foi causing the pioblem
DO! ?)%$&-&=/$&)* - uefineu as when the accuseu paity aumits iesponsibility
foi a ceitain event anu asks foigiveness fiom the geneial public anu
those impacteu
*,C D! M)8&$&N,÷favois the actions taken ¡ comments maue by BP
E! Ne0ative÷ciiticizes the actions taken ¡ comments maue by BP
L! Neutial÷neithei favois noi ciiticizes the actions taken ¡ comments maue
by BP oi theie aie a seemingly equal amount of positive anu negative
comments
9,*:,% )- 8,*:,%C
D! Female
E! Nale
L! 0thei
49
?,88/0, <)=(8C
D! Inuustiy
E! Bealth
L! Enviionment
O! Financial anu economic impact
P! Recoveiy
Q! 0thei
+/$, )- ?,88/0,
@*8$%(=$&)*8C
Coueis aie iesponsible foi accuiately categoiizing all uata gatheieu as
coiiectly as possible. Fiist, coueis enteieu the uate that the unit of analysis was
wiitten in a NN¡BB¡YY foimat. Seconu, when applicable, coueis categoiizeu the
wiitei of the unit being analyzeu as Nale, Female, oi 0thei. Because the content
being analyzeu was coming fiom social meuia outlets, theie may be some names
that aie aliases oi not ieal names. These weie categoiizeu unuei "0thei," as well as
any names that aie unisex oi otheiwise not cleai. Names that aie tiauitionally
women's names weie categoiizeu as "Female" anu tiauitionally masculine names as
"Nale."
All uata in the sample geneiateu fiom BP was collecteu fiom the company's
veiifieu Twittei account. These tweets weie coueu foi uate, iestoiation stiategy,
anu when applicable, the paiticulai Image Restoiation Stiategy. To coue content foi
Image Restoiation Stiategy (Benoit, 1997), coueis iuentifieu the stiategy BP
Su
employeu in cieating the message content. Coueis weie given a list of the 14
stiategies anu examples of each befoie the couing piocess begins.
The uata set that is composeu of comments geneiateu by the public uiiecteu
towaius BP was then coueu to iuentify tone, iefeience to paiticulai subject aiea,
uate, anu genuei of wiitei. Content that was wiitten in suppoit of BP oi theii effoits
to clean the uulf, iectify the situation, oi take iesponsibility foi what they have uone
was categoiizeu as "Positive." Also to be categoiizeu as positive weie comments
about measuies that aie peiceiveu as effective oi comments about the
impiovements that BP is iesponsible foi in the wake of the oil spill.
Content that conuemneu BP's posteu message, theii measuies in the uulf,
anu theii iesponse to the oil spill in geneial weie categoiizeu as "Negative." Content
that ciiticizeu BP's iesponse as ineffective, insufficient, oi impiopei was categoiizeu
as negative. Content that challenges BP's honesty anu integiity in the cleanup oi
afteimath in any way was also categoiizeu as negative.
Some content hau both positive anu negative iemaiks towaiu BP anu theii
iesponse to the oil spill. Content that incluueu equal paits negative anu positive was
categoiizeu as "Neutial." Also to be categoiizeu as neutial, weie those that geneially
expiesseu no stiong feeling in eithei uiiection about the actions being unueitaken.
Foi example, some iesponses might give a suggestion oi claiification, but weie
neithei ciitical noi wiitten in piaise. This content was categoiizeu as neutial.
The content in this sample set was coueu to ueteimine the paiticulai focus of
the message, incluuing: inuustiy, health, enviionment, financial anu economic
impact, iecoveiy, oi othei.
S1
Befoie the couing anu categoiization piocess began, coueis uiscusseu anu
came to agieement on the meaning of the categoiies above. 0nce agieement anu
unueistanuing was ieacheu, couing begin. Inteicouei ieliability will be testeu to
ensuie that a high level of agieement is ieacheu.
S2
APPENBIX B
Scieen shot of Tweets fiom ÇBP_Ameiica
SS
APPENBIX C
Scieen shot of Tweets ÇBP_Ameiica
doc_434314405.pdf