Description
Everyone is aware of the growth of digital communication about brands, including famous examples that illustrate the potential for major changes in consumer attitudes to be driven entirely though the internet. But few people know the typical levels of activity and influence, behind all the attention focused on a few high-profile examples.
AHEAD OF WHAT’S NEXT.
The Infuence of Digital Communication
on Consumers’ Attitudes to Brands
Kevin Ford, Joint Global Head of Brand and Communications Research,
& Andrew Freeman, Senior Consultant Media Research
1
Overview
Everyone is aware of the growth of digital communication about brands, including famous
examples that illustrate the potential for major changes in consumer attitudes to be driven
entirely though the internet. But few people know the typical levels of activity and infuence,
behind all the attention focused on a few high-profle examples.
This paper draws on syndicated and custom studies by Harris Interactive, that have enabled
us to study diferent types of communications across hundreds of UK brands. This includes
measuring the extent of the communications, their strength of overall infuence and their
difering roles as infuencers of our hearts and minds.
The big picture is that, on average, digital communications are less prevalent but more
infuential than traditional forms of communication. Put another way, traditional marketing
still has a wider reach across the population, but typically does not carry the same explosive
potential for change that characterises the most extreme examples in the digital space.
Digital marketing also plays a slightly diferent role, emphasising communication of more
functional, practical information than non-digital media, particularly in comparison to
television advertising.
This general pattern is also true and indeed more marked for word-of-mouth, including
any communications not directly managed by the brand owner. Their incidence is small but
their infuence is large, particularly so for digital word-of-mouth.
Putting it all together, this paper illustrates the difering roles and potential for each type
of communication, including the risks of negative reactions as well as the potential for
positive results. It demonstrates that the most efective course of action is typically an
integrated management of communications, avoiding the temptation to concentrate
on just one medium, and employing the diferent types in the most synergistic way.
We therefore hope it will guide brand managers to a better understanding and management
of their target audience and communications.
2
The Infuence of Digital Communication...
The Surveys
The main study this paper draws on is the EquiTrend survey, and it is worth spending
a little time on what this study is and how it has evolved.
The survey has been run by Harris Interactive in USA for over a decade, covering brand equity
and consumer connection for over 1000 brands. The study is syndicated and fndings from it
are released to companies individually, to help them benchmark their brands’ performance
against a spectrum of direct competitors and the broader range of great performers and
underachievers.
The EquiTrend study was also run recently in the UK, interviewing over 4000 people in
September 2009. Interviewing was conducted on the Harris Interactive Online Access Panel,
weighted to be representative of the total adult population. The survey included 243 brands
across 20 categories, with the same survey content as previous USA studies, plus a focus on
consumers’ awareness or experience of marketing and other touchpoints with each brand.
There was also a particular focus on digital activity, by which we mean:
Advertising on a website, a search engine or instant messenger, an email or
•
text message
Visits to the brand’s own website
•
Visibility of the brand on retail or information websites, such as price comparison sites
•
‘Inbound’ digital word-of-mouth, e.g. seeing comments about the brand via a blog,
•
chat room, YouTube or social network
‘Outbound’ digital word-of-mouth, e.g. creating word-of-mouth through a blog,
•
chat room, etc
We used all this data to study general patterns of infuence across brands and categories.
Analysis focused on people who had used each type of category in the past three months,
and were familiar with each brand. All our measurements and conclusions are therefore
derived from a relevant target audience.
In addition to the EquiTrend survey, we also run a great many custom studies for individual
clients. These provide extra material and the opportunity to go into specifc cases in greater
depth. Findings from a couple of these are mentioned in this paper, with some details have
been disguised for reasons of confdentiality.
3
The Infuence of Digital Communication...
How Much Activity is There?
You might believe from the pronouncements of some commentators that digital
communications about brands is very prevalent, reaching the majority of the population.
However, this is not the case.
Figure 1 shows the profle of the communications ‘surround’ for an average brand, listing
the percentage of people claiming to have seen or heard each type of communication
in the past month for any specifc band. The data reveals that relatively few people are
exposed to digital communication. Only 12% of people had seen any form of digital
communication, contrasting with as much as half the population claiming a non-digital
experience of that brand. And these diferences carry through to the balance for specifc
types of communication:
Marketing: digital 11%, non-digital 40%
•
Inbound word-of-mouth: digital 2%, non-digital 5%
•
Outbound word-of-mouth: digital 1%, non-digital 3%
•
Of course, it must be stressed that these fgures represent the numbers of people, not the
quantity or events or size of infuence among those people. They are simply a salient reminder
that digital communications do not reach large numbers of people for the vast majority of
brands. Exceptions exist, but this is the general picture.
12% of category users
familiar with the brand had
some form of digital
experience with it in the
past month
Told others about it
Word-of-mouth
(seen/heard on a blog, chat room, Youtube, social network)
Retail or info website
Brand’s website
Marketing (adverts, emails)
Any Digital Experience
Told others about it
Word-of-mouth
(from a friend, in the news, ratings magazine)
Seen out and about
Purchase or service experience
Marketing
(TV, Press, radio, poster, sponsorship, promotions, direct mail)
Any Non–Digital Experience
1%
2%
3%
5%
11%
12%
3%
5%
12%
15%
40%
52%
Non-Digital Surround
Figure 1: The Profle of the ‘Surround’ for an average brand
Digital Surround
Continuing the story, we looked at the level of digital activity across the 20 diferent
categories covered by the EquiTrend survey, and found a wide spread of fgures.
4
The Infuence of Digital Communication...
On the one hand, at the top are the airlines – and this will be driven in part by seasonality
(remember we carried out the survey during the summer, when the airlines would all
have been pushing themselves very heavily). There are also the traditional suspects of
technology and telecoms, along with the myriad entertainment brands that have long
adopted the web.
At the bottom are categories like petrol, food and drink, and restaurants. These categories
all have slightly diferent challenges. For restaurants it will be accurate localisation, and
the challenge of knowing when to advertise – a challenge which search optimisation turns
into a genuine opportunity. Reach will tend to remain low, but the people these digital
touchpoints do reach are more likely to be in the mood for that impulse visit. For Food and
Drink brands, traditional retail and marketing channels are the key media for most brands,
most of the time. Note, however, that supermarket shopping sites are high in this list.
Figure 2: Airlines, entertainment and services have the most digital activity
Airlines
Entertainment
Telecoms
Technology
Insurance companies
Retail – supermarkets
Gambling
Credit cards
Media (TV, Radio, News/Publishing
Retail – larger high street stores
Utilities
Banks and building societies
Household appliances
Retail – clothing
Auto
Personal care
Restaurants
Food
Drink
Petrol
For an average airline brand,
27% of people had
experienced some form of
digital activity
0% 5% 10% 15% 20% 25% 30%
And fnally petrol: the huge oil company brands obviously have an online presence, but
perhaps for brands like this we still need a little more development of the mobile digital
space, and the gradual digitisation of the car, before we’ll see these low-interest, but essential
brands begin to connect more with people’s digital lives. There is already a TomTom app for
the iPhone, so this revolution has begun – we look forward to seeing how it develops.
5
The Infuence of Digital Communication...
Delving deeper, we looked at individual brands, identifying those at the high and low
ends of the spectrum of levels of activity. Figure 3 (over) lists the players for digital activity
as a whole, while fgure 4 pulls out those which are particularly the subject of digital
word-of-mouth.
The most interesting aspect of these two lists is the relatively low overlap between them.
We have highlighted in fgure 4 those brands which also appeared in the equivalent lists in
fgure 3. It shows that only the very few brands highlighted are both overall most active and
most talked about, the specifc cases being YouTube, iPhone and Microsoft. This is important,
because it should remind you that the talkability of your brand is not a function of your own
investment – at least not a direct consequence. Once your content and your brand is in the
digital space, will people want to talk about it? Do you empower them to talk about it?
It is no surprise to see You Tube or the iPhone among the most talked about list – they have
succeeded in turning their signifcant online presence into something that people want to
comment on or talk about. For brands like Xbox and New Look, they have done so much
more efectively – they are not so active overall in the digital space, but where they are they
have been much more successful in turning this lower level of activity into social buzz.
The Headliners
(top 15)
British Airways
Legal & General
YouTube
Tesco
iTunes
Microsoft
iPhone
Apple
Sky
Boots
Argos
More Th>n
Ladbrokes
Virgin Credit Cards
Ryan Air
The Quiet Ones
(bottom 15)
KFC
Black & Decker
Esso
H&M
Total
La Tasca
MTV
Focus DIY
Q8
KitKat
Soltan
Co-op
Twinings Tea
TGIs
Daily Express
Note: some brands omitted due to small base size
Figure 3: Who has the most digital
activity?
Figure 4: Who gets talked about on
the web?
The Headliners
(top 15)
YouTube
Xbox 360
iPhone
New Look
Seeboard Energy
MBNA
Skype
Microsoft
3 mobile
Virgin Mobile
Absolute Radio
Peugeot
Blackberry
EA Games
Ladbrokes
The Quiet Ones
(bottom 15)
Nivea
Daily Express
TGIs
Oasis (soft drink)
Total
The Financial Times
Esso
HSBC
Matalan
MINT
Canon
Acer
Mercedes
Laughing Cow
Douwe Egberts
Note: some brands omitted due to small base size
6
The Infuence of Digital Communication...
The Strength of Infuence on Brand Attitudes and Outcomes
It is one thing to count the number of people engaged in diferent brand experiences
and communications. A much bigger challenge is to measure the power and nature
of the infuence of those experiences.
First, we need to deal with a problem called the Rosser Reeves fallacy, the tendency
to confuse cause and efect in advertising research. In essence, advertising research
so often seems to work because it shows that people who noticed the advertising are
more likely to like your brand. What this ignores is the fact that people who already
like your brand are more likely to notice your advertising. And this is exacerbated in a
multi-media world, where noticing a campaign in one medium makes you more sensitive to
it in other guises.
So our modelling of the Touchpoints involves weighting respondent data to account
for the existing afnity with or use of the brand, by balancing the profles of users and non-
users, both against each other, and against the profles of diferent media recall
within them.
As an example of this, fgure 5 shows the efect of the modelling on a specifc custom
study. The data shown are real, but we have disguised the specifc attributes in order
to protect confdentiality. On the left are bars showing the levels of agreement with
a series of brand attitude statements, comparing people who remember seeing the
advertising (‘experienced’) against those who did not (‘not experienced’). As you
can see, there will be a marketing manager somewhere with a great big grin, because
everything is up. But this goes against what we know about marketing, and what we
know about this campaign: this campaign did not talk about all these diferent aspects
of the brand.
Figure 5: An example of Touchpoint modelling…
Unadjusted Data
Uplifts are infated – ranging from +9 to +17%
Category generics are strongest performers
Adjusted Data
Uplifts are corrected– ranging from 0 to +7%
Key campaign messages are strongest performers
Attribute 7
Attribute 6
Attribute 5
Attribute 4
Attribute 3
Attribute 2
Attribute 1
16%
13%
23%
18%
15%
25%
17%
25%
30%
32%
23%
25%
28%
31%
14%
12%
28%
21%
14%
22%
27%
37%
27%
44%
21%
23%
34%
36%
Not experienced Experienced
7
The Infuence of Digital Communication...
On the right we’ve applied the touchpoint modelling, and the scale of diferences is much
more reasonable: only three are left with signifcant changes, and these are the ones that the
campaign did focus on. The modelling also refned the story – we now see that attribute 1
was more afected than attribute 3 by the advertising, which was not visible in the raw data.
So we can say that yes, the campaign was a success, and yes the marketing manager should
be happy: but he cannot plan his next campaign on the basis that whatever he says,
everything will go up.
So, using this Touchpoint Modelling method we have an approach that achieves two
main things:
Size of infuence: It correctly measures the size of the infuence from a communication
•
channel
Attribution of infuence: it correctly attributes infuences to the right medium, since it
•
controls for people’s exposure to all other media.
This therefore gives us a very powerful tool for examining the efects of diferent
communications, in a world where people are immersed in a ‘surround’ of multiple infuences.
Our touchpoint modelling has been applied to all the analyses in the rest of this paper.
Returning to the EquiTrend survey, fgure 6 profles the infuence of fve major types of
communications on people’s likelihood to recommend (say good things) or denounce (say
bad things) about the brand.
As a reminder, touchpoint modelling has been used to ensure that we correctly measure the
infuence of each communication, avoiding misattribution between them.
Figure 6: Among those exposed to it, word of mouth is the big infuencer,
particularly digital word of mouth
6%
2%
9%
0%
TV Other nondig
mktg
WOM
non-dig
WOM
Digital
Mktg/ website
Digital
25%
20%
15%
10%
5%
4%
14%
20%
14%
8%
9%
4%
Exposure to digital
word-of-mouth increases
likelihood to recommend to
others by 20%
Recommend often
Denounce often
Uplifts in recommendation and denouncement, among those exposed to each touchpoint
(data modelled using Harris Interactive touchpoint modelling).
8
The Infuence of Digital Communication...
Three very important themes are evident.
1. Digital experiences have a more powerful efect than non-digital experiences. For example,
if we compare digital versus non-digital word-of-mouth, we see that the digital version
has a greater infuence on recommendation (+20%). Similarly, digital marketing creates a
stronger reaction than of-line marketing.
2. Word-of mouth has a more powerful efect than managed communication. This is visible
in both the digital and non-digital forms.
3. All channels carry some risk of negative outcomes as well as positive outcomes.
The infuence on recommendation is higher in all cases, but the infuence on denouncement
is also noticeable, running at about a third to a half of the level of recommendation.
One lesson from this is that it is easy to pass on word-of-mouth, particularly in the Digital
world. It means that brands need to provide easily forwardable content and connections
to beneft from that, and equally need to address negative sentiment fast.
While tendency to pass on comment gives us some steer, a more powerful test is given by
the efect on brand consideration (see fgure 7). This focuses on the end result of all this
recommendation, denouncement and experience in general.
We asked our sample to what extent they would consider each brand. Was it…
the only brand they would consider
•
one of a few they would consider
•
one of many they might consider (in efect this was the neutral, undecided option)
•
a brand I would never consider
•
3%
1%
-7%
4%
3%
4%
-12%
6%
4%
-2%
-12%
10%
10%
-7%
-11%
8%
6%
2%
-12%
3%
15%
10%
5%
0%
-5%
-10%
-15%
TV Other nondig
mktg
WOM non-dig WOM Digital MKTG/ website
Digital
The only one I would consider
One of many brands I might consider
One of a few brands I would consider
A brand I would never consider
In person, we tell good news to
one or two people, but bad
news to many more
Digital word-of-mouth
increases positive preference
and rejection
Figure 7: Marketing and word-of-mouth help you make up your
mind, positively or negatively
9
The Infuence of Digital Communication...
The data in fgure 7 shows the change in the percentage giving each response, between
those exposed to the communication versus those not exposed. For example, those exposed
to TV advertising were 3% more likely to give the top response: the only brand I would
consider (look at the far left of the chart).
Here again, we see a number of important fndings.
1. Any communication helps you make up your mind. For all fve types of communication,
the percentage giving the ‘neutral/undecided’ response drops after exposure to the
communication.
2. The efect can be positive or negative. All forms of communication carry the risk of negative
infuence as well as positive infuence. This makes intuitive sense for marketing
– once you have heard the sales pitch you can pretty quickly work out whether this is right
for you.
3. The spoken word carries a particular risk of a negative outcome. We noted on fgure 6
that more people were recommending than denouncing the brand as a consequence of
non-digital word-of-mouth, but fgure 7 adds in some extra dimensions. In person, we tell
good news to one or two people but bad news to many more. And the horror stories have
more ‘sticking power’ than the nice stories.
4. Social media and blogs increase both positive preference and brand rejection. This
research shows that monitoring social media simply as a disaster-monitor is missing the
real power of digital marketing. People who pass on digital content and comment are
investing at least some time in your brand, and in general, that is because they judge your
brand to be worth spending time on. Companies that spend time and efort creating
content and comment that is easy and quick to pass on will fnd plenty of willing volunteers
to do so, and this has a huge impact on brand consideration.
5. Digital marketing produces more positive results than traditional marketing. The infuence
fgures for traditional marketing are essentially balanced: equal increases in both positive
consideration and rejection of the brand. However, the fgures for digital marketing are
more clearly positive, showing an 8% increase in the considerers, compared to only a 3%
increase in the rejecters. This perhaps highlights that it is easier to ignore digital marketing
that is not relevant to you – you are unlikely to visit the brand’s own website unless the
brand already has some relevance for you, and it is easy to avoid engaging with digital
advertising that covers only part of a screen. In efect, we choose what we engage with in
the digital space, more than with traditional media. There can also appear to be an honesty
about the digital world, because it appears, or can be made to appear, less controlled
by the brand owners. For example, many brands’ websites allow user reviews to be posted
on them.
A pattern therefore emerges. The communication channels and types that have the lower
levels of reach across the population tend to be the ones that have the greater infuence. This
includes anything digital, and in particular word-of-mouth on the web. Putting those two
things together we can calculate the total efect, i.e. the reach times the reaction.
10
The Infuence of Digital Communication...
Figure 8 shows the result for the three aspects for each of type of communication:
the percentage reached (vertical axis),
•
the reaction, i.e. the uplift in top box consideration (horizontal axis), and
•
the percentage of the ‘total efect’ accounted for by that communication
•
(shown in brackets).
All three types of marketing account for large proportions of the total efect: TV 35%, other
non-digital marketing 24% and digital marketing 27%. Allowing for the likelihood of some
misattribution of awareness in favour of TV, we might consider all of these fgures to be
broadly similar. This aligns with the recent fgures from the IAB that show online spend
overtaking TV. By contrast, the total infuence of all word-of-mouth is noticeably lower for an
average brand, since although it has high impact per person touched, it has a much lower
reach across the population.
Figure 8: Digital Marketing has a strong overall infuence on brand consideration
0%
Uplift on top box consideration
35%
2% 4% 6% 8% 10% 12% 0%
30%
25%
15%
10%
20%
5%
TV
(35%)
Mktg/website Digital
(27%)
WOM Digital
(6%)
WOM non-dig
(9%)
Other nondig mktg
(24%)
The net efect of digital marketing is similar to
TV or other non-digital marketing. It reaches fewer
people but has a stronger efect on those it touches
11
The Infuence of Digital Communication...
Using the Media in Combination
The above analysis carries the risk that a brand manager may conclude that the right course
of action is to focus all their energies on the most powerful medium. But this ignores two
important facts, demonstrated in the remainder of this paper.
1. Using a combination of media increases the efect
2. Diferent media create diferent efects
First let’s look at the potential to increase the efect by using multiple media. Figure 9 looks at
the levels of recommendation among diferent subgroups of people, categorised by their
exposure to specifc combinations of marketing. To make the analysis clear and focused, for
this particular chart we have looked only at current users of the brand, and considered
combinations of types of marketing. Word of mouth has been controlled to be equivalent for
each subgroup in the chart.
The results are extreme. Levels of recommendation range from only 13% among people who
have seen no marketing, through to 40% among people exposed to all three media. And
there is a subtlety in the fgures in between: there is no visible ‘tail-of’ in efect as we increase
the number of media, and the combination of digital and non-digital is more benefcial than
multiple types of traditional media.
Figure 9: Multiple media create the maximum efect
% recommending to others, within each group
All three
Other nondigital mktg AND
Digital mkg/website
TV AND digital mktg/website
TV AND other nondigital mktg
Digital mktg/website only
Other nondigital mktg only
TV only
None
0% 10% 20% 30% 40% 50%
Three
types
Two
types
One
type
None
Exposure to more media
increases likelihood to
recommend the brand
40%
34%
31%
27%
25%
24%
21%
13%
The analysis focuses on recommendation by brand users, for all
combinations of three types of media
12
The Infuence of Digital Communication...
This makes sense. Many advertisers deliberately reference the internet in their traditional
communications, inviting people to visit their website. Traditional media frst engage the
audience, then digital activity helps people along the pathway to purchase, or simply builds
on an initial level of interest. Prime examples of this efect would be campaigns like Cadbury’s
Gorilla or the Sony Bravia bouncing balls. The lesson from this analysis is to use the media in a
combination best suited to your brand and category, rather than put all your eggs into a
single basket.
Finally, again drawing from a diferent study that also used our touchpoint modelling, here is
a further reason why it is too early to proclaim the death of TV.
This is taken from a study we recently carried out in which our client had a campaign running
across a variety of media, from TV and Press creative, to online ads and Point of Sale material.
Everyone we spoke to in this study was in the market for the product – but when we look at
the infuences after running our touchpoint modelling, we can identify that those people
exposed to the TV work were much more emotionally impacted, while those exposed to the
online and point of sale material had bigger changes to their rational connections with the
brand.
Validation work from our EquiTrend study shows that for the vast majority of categories
it is important to build both an emotional and rational connection with the brand, in order to
attract and sustain a relationship with a customer. And in many categories, including
this example, it is necessary to stimulate the consumer to consider the whole idea of making
the purchase.
In this case, online advertising used on its own would have run the risk of reaching only a narrow
target audience, who were already somewhat pre-disposed to engage with the brand and seek
information. Traditional media used on its own, in particular TV, would have run the opposite
risk of awakening some interest but failing to follow through and ‘close the sale’.
Not shown in fgure 10, but identifed in the EquiTrend study was that word-of-mouth tends
to have a strong emotional reaction, understandably so given the tone of expression
employed in many word-of-mouth communications!
Figure 10: Choosing between media means not
choosing efect, but choosing between type of efect
TV
Press
Internet
POS
31%
40%
46%
59%
69%
60%
54%
41%
Rational Emotional
AHEAD OF WHAT’S NEXT.
The Infuence of Digital Communication...
Summary and Recommendations
In conclusion, let’s draw together these thoughts.
For all the rhetoric about online and social media, the ability of brands to dominate across a
medium with four billion pages is currently very limited; the reach of a brand’s campaign in
digital media will remain small, or fragmented, but its potential infuence among that group
could be large. Online therefore ofers the ‘audience of interest’ that traditional broadcast
fails to deliver efciently.
However, online works best when it is used to sharpen a message – a message that traditional
broadcast media are already carrying. While digital marketing ofers those users all the
information they need, traditional techniques ofer the emotional connection that is needed
to spark that need. Traditional media builds consideration: I’ve heard of that brand, and it is
worthy of my click.
In a world where social media is increasingly important, digital word-of-mouth is the epitome
of digital marketing: still very hard for a brand to achieve meaningful reach amidst the billions
of conversations, but intensely powerful when it does so. And that power needs to be
monitored, and indeed fed, (with content or comment) if brands are to properly harness it.
AHEAD OF WHAT’S NEXT.
© 2009 Harris Interactive, Inc. All Rights Reserved. 10.09
+44 (0) 20 8263 5200 [email protected] www.harrisinteractive.com/europe
doc_191314516.pdf
Everyone is aware of the growth of digital communication about brands, including famous examples that illustrate the potential for major changes in consumer attitudes to be driven entirely though the internet. But few people know the typical levels of activity and influence, behind all the attention focused on a few high-profile examples.
AHEAD OF WHAT’S NEXT.
The Infuence of Digital Communication
on Consumers’ Attitudes to Brands
Kevin Ford, Joint Global Head of Brand and Communications Research,
& Andrew Freeman, Senior Consultant Media Research
1
Overview
Everyone is aware of the growth of digital communication about brands, including famous
examples that illustrate the potential for major changes in consumer attitudes to be driven
entirely though the internet. But few people know the typical levels of activity and infuence,
behind all the attention focused on a few high-profle examples.
This paper draws on syndicated and custom studies by Harris Interactive, that have enabled
us to study diferent types of communications across hundreds of UK brands. This includes
measuring the extent of the communications, their strength of overall infuence and their
difering roles as infuencers of our hearts and minds.
The big picture is that, on average, digital communications are less prevalent but more
infuential than traditional forms of communication. Put another way, traditional marketing
still has a wider reach across the population, but typically does not carry the same explosive
potential for change that characterises the most extreme examples in the digital space.
Digital marketing also plays a slightly diferent role, emphasising communication of more
functional, practical information than non-digital media, particularly in comparison to
television advertising.
This general pattern is also true and indeed more marked for word-of-mouth, including
any communications not directly managed by the brand owner. Their incidence is small but
their infuence is large, particularly so for digital word-of-mouth.
Putting it all together, this paper illustrates the difering roles and potential for each type
of communication, including the risks of negative reactions as well as the potential for
positive results. It demonstrates that the most efective course of action is typically an
integrated management of communications, avoiding the temptation to concentrate
on just one medium, and employing the diferent types in the most synergistic way.
We therefore hope it will guide brand managers to a better understanding and management
of their target audience and communications.
2
The Infuence of Digital Communication...
The Surveys
The main study this paper draws on is the EquiTrend survey, and it is worth spending
a little time on what this study is and how it has evolved.
The survey has been run by Harris Interactive in USA for over a decade, covering brand equity
and consumer connection for over 1000 brands. The study is syndicated and fndings from it
are released to companies individually, to help them benchmark their brands’ performance
against a spectrum of direct competitors and the broader range of great performers and
underachievers.
The EquiTrend study was also run recently in the UK, interviewing over 4000 people in
September 2009. Interviewing was conducted on the Harris Interactive Online Access Panel,
weighted to be representative of the total adult population. The survey included 243 brands
across 20 categories, with the same survey content as previous USA studies, plus a focus on
consumers’ awareness or experience of marketing and other touchpoints with each brand.
There was also a particular focus on digital activity, by which we mean:
Advertising on a website, a search engine or instant messenger, an email or
•
text message
Visits to the brand’s own website
•
Visibility of the brand on retail or information websites, such as price comparison sites
•
‘Inbound’ digital word-of-mouth, e.g. seeing comments about the brand via a blog,
•
chat room, YouTube or social network
‘Outbound’ digital word-of-mouth, e.g. creating word-of-mouth through a blog,
•
chat room, etc
We used all this data to study general patterns of infuence across brands and categories.
Analysis focused on people who had used each type of category in the past three months,
and were familiar with each brand. All our measurements and conclusions are therefore
derived from a relevant target audience.
In addition to the EquiTrend survey, we also run a great many custom studies for individual
clients. These provide extra material and the opportunity to go into specifc cases in greater
depth. Findings from a couple of these are mentioned in this paper, with some details have
been disguised for reasons of confdentiality.
3
The Infuence of Digital Communication...
How Much Activity is There?
You might believe from the pronouncements of some commentators that digital
communications about brands is very prevalent, reaching the majority of the population.
However, this is not the case.
Figure 1 shows the profle of the communications ‘surround’ for an average brand, listing
the percentage of people claiming to have seen or heard each type of communication
in the past month for any specifc band. The data reveals that relatively few people are
exposed to digital communication. Only 12% of people had seen any form of digital
communication, contrasting with as much as half the population claiming a non-digital
experience of that brand. And these diferences carry through to the balance for specifc
types of communication:
Marketing: digital 11%, non-digital 40%
•
Inbound word-of-mouth: digital 2%, non-digital 5%
•
Outbound word-of-mouth: digital 1%, non-digital 3%
•
Of course, it must be stressed that these fgures represent the numbers of people, not the
quantity or events or size of infuence among those people. They are simply a salient reminder
that digital communications do not reach large numbers of people for the vast majority of
brands. Exceptions exist, but this is the general picture.
12% of category users
familiar with the brand had
some form of digital
experience with it in the
past month
Told others about it
Word-of-mouth
(seen/heard on a blog, chat room, Youtube, social network)
Retail or info website
Brand’s website
Marketing (adverts, emails)
Any Digital Experience
Told others about it
Word-of-mouth
(from a friend, in the news, ratings magazine)
Seen out and about
Purchase or service experience
Marketing
(TV, Press, radio, poster, sponsorship, promotions, direct mail)
Any Non–Digital Experience
1%
2%
3%
5%
11%
12%
3%
5%
12%
15%
40%
52%
Non-Digital Surround
Figure 1: The Profle of the ‘Surround’ for an average brand
Digital Surround
Continuing the story, we looked at the level of digital activity across the 20 diferent
categories covered by the EquiTrend survey, and found a wide spread of fgures.
4
The Infuence of Digital Communication...
On the one hand, at the top are the airlines – and this will be driven in part by seasonality
(remember we carried out the survey during the summer, when the airlines would all
have been pushing themselves very heavily). There are also the traditional suspects of
technology and telecoms, along with the myriad entertainment brands that have long
adopted the web.
At the bottom are categories like petrol, food and drink, and restaurants. These categories
all have slightly diferent challenges. For restaurants it will be accurate localisation, and
the challenge of knowing when to advertise – a challenge which search optimisation turns
into a genuine opportunity. Reach will tend to remain low, but the people these digital
touchpoints do reach are more likely to be in the mood for that impulse visit. For Food and
Drink brands, traditional retail and marketing channels are the key media for most brands,
most of the time. Note, however, that supermarket shopping sites are high in this list.
Figure 2: Airlines, entertainment and services have the most digital activity
Airlines
Entertainment
Telecoms
Technology
Insurance companies
Retail – supermarkets
Gambling
Credit cards
Media (TV, Radio, News/Publishing
Retail – larger high street stores
Utilities
Banks and building societies
Household appliances
Retail – clothing
Auto
Personal care
Restaurants
Food
Drink
Petrol
For an average airline brand,
27% of people had
experienced some form of
digital activity
0% 5% 10% 15% 20% 25% 30%
And fnally petrol: the huge oil company brands obviously have an online presence, but
perhaps for brands like this we still need a little more development of the mobile digital
space, and the gradual digitisation of the car, before we’ll see these low-interest, but essential
brands begin to connect more with people’s digital lives. There is already a TomTom app for
the iPhone, so this revolution has begun – we look forward to seeing how it develops.
5
The Infuence of Digital Communication...
Delving deeper, we looked at individual brands, identifying those at the high and low
ends of the spectrum of levels of activity. Figure 3 (over) lists the players for digital activity
as a whole, while fgure 4 pulls out those which are particularly the subject of digital
word-of-mouth.
The most interesting aspect of these two lists is the relatively low overlap between them.
We have highlighted in fgure 4 those brands which also appeared in the equivalent lists in
fgure 3. It shows that only the very few brands highlighted are both overall most active and
most talked about, the specifc cases being YouTube, iPhone and Microsoft. This is important,
because it should remind you that the talkability of your brand is not a function of your own
investment – at least not a direct consequence. Once your content and your brand is in the
digital space, will people want to talk about it? Do you empower them to talk about it?
It is no surprise to see You Tube or the iPhone among the most talked about list – they have
succeeded in turning their signifcant online presence into something that people want to
comment on or talk about. For brands like Xbox and New Look, they have done so much
more efectively – they are not so active overall in the digital space, but where they are they
have been much more successful in turning this lower level of activity into social buzz.
The Headliners
(top 15)
British Airways
Legal & General
YouTube
Tesco
iTunes
Microsoft
iPhone
Apple
Sky
Boots
Argos
More Th>n
Ladbrokes
Virgin Credit Cards
Ryan Air
The Quiet Ones
(bottom 15)
KFC
Black & Decker
Esso
H&M
Total
La Tasca
MTV
Focus DIY
Q8
KitKat
Soltan
Co-op
Twinings Tea
TGIs
Daily Express
Note: some brands omitted due to small base size
Figure 3: Who has the most digital
activity?
Figure 4: Who gets talked about on
the web?
The Headliners
(top 15)
YouTube
Xbox 360
iPhone
New Look
Seeboard Energy
MBNA
Skype
Microsoft
3 mobile
Virgin Mobile
Absolute Radio
Peugeot
Blackberry
EA Games
Ladbrokes
The Quiet Ones
(bottom 15)
Nivea
Daily Express
TGIs
Oasis (soft drink)
Total
The Financial Times
Esso
HSBC
Matalan
MINT
Canon
Acer
Mercedes
Laughing Cow
Douwe Egberts
Note: some brands omitted due to small base size
6
The Infuence of Digital Communication...
The Strength of Infuence on Brand Attitudes and Outcomes
It is one thing to count the number of people engaged in diferent brand experiences
and communications. A much bigger challenge is to measure the power and nature
of the infuence of those experiences.
First, we need to deal with a problem called the Rosser Reeves fallacy, the tendency
to confuse cause and efect in advertising research. In essence, advertising research
so often seems to work because it shows that people who noticed the advertising are
more likely to like your brand. What this ignores is the fact that people who already
like your brand are more likely to notice your advertising. And this is exacerbated in a
multi-media world, where noticing a campaign in one medium makes you more sensitive to
it in other guises.
So our modelling of the Touchpoints involves weighting respondent data to account
for the existing afnity with or use of the brand, by balancing the profles of users and non-
users, both against each other, and against the profles of diferent media recall
within them.
As an example of this, fgure 5 shows the efect of the modelling on a specifc custom
study. The data shown are real, but we have disguised the specifc attributes in order
to protect confdentiality. On the left are bars showing the levels of agreement with
a series of brand attitude statements, comparing people who remember seeing the
advertising (‘experienced’) against those who did not (‘not experienced’). As you
can see, there will be a marketing manager somewhere with a great big grin, because
everything is up. But this goes against what we know about marketing, and what we
know about this campaign: this campaign did not talk about all these diferent aspects
of the brand.
Figure 5: An example of Touchpoint modelling…
Unadjusted Data
Uplifts are infated – ranging from +9 to +17%
Category generics are strongest performers
Adjusted Data
Uplifts are corrected– ranging from 0 to +7%
Key campaign messages are strongest performers
Attribute 7
Attribute 6
Attribute 5
Attribute 4
Attribute 3
Attribute 2
Attribute 1
16%
13%
23%
18%
15%
25%
17%
25%
30%
32%
23%
25%
28%
31%
14%
12%
28%
21%
14%
22%
27%
37%
27%
44%
21%
23%
34%
36%
Not experienced Experienced
7
The Infuence of Digital Communication...
On the right we’ve applied the touchpoint modelling, and the scale of diferences is much
more reasonable: only three are left with signifcant changes, and these are the ones that the
campaign did focus on. The modelling also refned the story – we now see that attribute 1
was more afected than attribute 3 by the advertising, which was not visible in the raw data.
So we can say that yes, the campaign was a success, and yes the marketing manager should
be happy: but he cannot plan his next campaign on the basis that whatever he says,
everything will go up.
So, using this Touchpoint Modelling method we have an approach that achieves two
main things:
Size of infuence: It correctly measures the size of the infuence from a communication
•
channel
Attribution of infuence: it correctly attributes infuences to the right medium, since it
•
controls for people’s exposure to all other media.
This therefore gives us a very powerful tool for examining the efects of diferent
communications, in a world where people are immersed in a ‘surround’ of multiple infuences.
Our touchpoint modelling has been applied to all the analyses in the rest of this paper.
Returning to the EquiTrend survey, fgure 6 profles the infuence of fve major types of
communications on people’s likelihood to recommend (say good things) or denounce (say
bad things) about the brand.
As a reminder, touchpoint modelling has been used to ensure that we correctly measure the
infuence of each communication, avoiding misattribution between them.
Figure 6: Among those exposed to it, word of mouth is the big infuencer,
particularly digital word of mouth
6%
2%
9%
0%
TV Other nondig
mktg
WOM
non-dig
WOM
Digital
Mktg/ website
Digital
25%
20%
15%
10%
5%
4%
14%
20%
14%
8%
9%
4%
Exposure to digital
word-of-mouth increases
likelihood to recommend to
others by 20%
Recommend often
Denounce often
Uplifts in recommendation and denouncement, among those exposed to each touchpoint
(data modelled using Harris Interactive touchpoint modelling).
8
The Infuence of Digital Communication...
Three very important themes are evident.
1. Digital experiences have a more powerful efect than non-digital experiences. For example,
if we compare digital versus non-digital word-of-mouth, we see that the digital version
has a greater infuence on recommendation (+20%). Similarly, digital marketing creates a
stronger reaction than of-line marketing.
2. Word-of mouth has a more powerful efect than managed communication. This is visible
in both the digital and non-digital forms.
3. All channels carry some risk of negative outcomes as well as positive outcomes.
The infuence on recommendation is higher in all cases, but the infuence on denouncement
is also noticeable, running at about a third to a half of the level of recommendation.
One lesson from this is that it is easy to pass on word-of-mouth, particularly in the Digital
world. It means that brands need to provide easily forwardable content and connections
to beneft from that, and equally need to address negative sentiment fast.
While tendency to pass on comment gives us some steer, a more powerful test is given by
the efect on brand consideration (see fgure 7). This focuses on the end result of all this
recommendation, denouncement and experience in general.
We asked our sample to what extent they would consider each brand. Was it…
the only brand they would consider
•
one of a few they would consider
•
one of many they might consider (in efect this was the neutral, undecided option)
•
a brand I would never consider
•
3%
1%
-7%
4%
3%
4%
-12%
6%
4%
-2%
-12%
10%
10%
-7%
-11%
8%
6%
2%
-12%
3%
15%
10%
5%
0%
-5%
-10%
-15%
TV Other nondig
mktg
WOM non-dig WOM Digital MKTG/ website
Digital
The only one I would consider
One of many brands I might consider
One of a few brands I would consider
A brand I would never consider
In person, we tell good news to
one or two people, but bad
news to many more
Digital word-of-mouth
increases positive preference
and rejection
Figure 7: Marketing and word-of-mouth help you make up your
mind, positively or negatively
9
The Infuence of Digital Communication...
The data in fgure 7 shows the change in the percentage giving each response, between
those exposed to the communication versus those not exposed. For example, those exposed
to TV advertising were 3% more likely to give the top response: the only brand I would
consider (look at the far left of the chart).
Here again, we see a number of important fndings.
1. Any communication helps you make up your mind. For all fve types of communication,
the percentage giving the ‘neutral/undecided’ response drops after exposure to the
communication.
2. The efect can be positive or negative. All forms of communication carry the risk of negative
infuence as well as positive infuence. This makes intuitive sense for marketing
– once you have heard the sales pitch you can pretty quickly work out whether this is right
for you.
3. The spoken word carries a particular risk of a negative outcome. We noted on fgure 6
that more people were recommending than denouncing the brand as a consequence of
non-digital word-of-mouth, but fgure 7 adds in some extra dimensions. In person, we tell
good news to one or two people but bad news to many more. And the horror stories have
more ‘sticking power’ than the nice stories.
4. Social media and blogs increase both positive preference and brand rejection. This
research shows that monitoring social media simply as a disaster-monitor is missing the
real power of digital marketing. People who pass on digital content and comment are
investing at least some time in your brand, and in general, that is because they judge your
brand to be worth spending time on. Companies that spend time and efort creating
content and comment that is easy and quick to pass on will fnd plenty of willing volunteers
to do so, and this has a huge impact on brand consideration.
5. Digital marketing produces more positive results than traditional marketing. The infuence
fgures for traditional marketing are essentially balanced: equal increases in both positive
consideration and rejection of the brand. However, the fgures for digital marketing are
more clearly positive, showing an 8% increase in the considerers, compared to only a 3%
increase in the rejecters. This perhaps highlights that it is easier to ignore digital marketing
that is not relevant to you – you are unlikely to visit the brand’s own website unless the
brand already has some relevance for you, and it is easy to avoid engaging with digital
advertising that covers only part of a screen. In efect, we choose what we engage with in
the digital space, more than with traditional media. There can also appear to be an honesty
about the digital world, because it appears, or can be made to appear, less controlled
by the brand owners. For example, many brands’ websites allow user reviews to be posted
on them.
A pattern therefore emerges. The communication channels and types that have the lower
levels of reach across the population tend to be the ones that have the greater infuence. This
includes anything digital, and in particular word-of-mouth on the web. Putting those two
things together we can calculate the total efect, i.e. the reach times the reaction.
10
The Infuence of Digital Communication...
Figure 8 shows the result for the three aspects for each of type of communication:
the percentage reached (vertical axis),
•
the reaction, i.e. the uplift in top box consideration (horizontal axis), and
•
the percentage of the ‘total efect’ accounted for by that communication
•
(shown in brackets).
All three types of marketing account for large proportions of the total efect: TV 35%, other
non-digital marketing 24% and digital marketing 27%. Allowing for the likelihood of some
misattribution of awareness in favour of TV, we might consider all of these fgures to be
broadly similar. This aligns with the recent fgures from the IAB that show online spend
overtaking TV. By contrast, the total infuence of all word-of-mouth is noticeably lower for an
average brand, since although it has high impact per person touched, it has a much lower
reach across the population.
Figure 8: Digital Marketing has a strong overall infuence on brand consideration
0%
Uplift on top box consideration
35%
2% 4% 6% 8% 10% 12% 0%
30%
25%
15%
10%
20%
5%
TV
(35%)
Mktg/website Digital
(27%)
WOM Digital
(6%)
WOM non-dig
(9%)
Other nondig mktg
(24%)
The net efect of digital marketing is similar to
TV or other non-digital marketing. It reaches fewer
people but has a stronger efect on those it touches
11
The Infuence of Digital Communication...
Using the Media in Combination
The above analysis carries the risk that a brand manager may conclude that the right course
of action is to focus all their energies on the most powerful medium. But this ignores two
important facts, demonstrated in the remainder of this paper.
1. Using a combination of media increases the efect
2. Diferent media create diferent efects
First let’s look at the potential to increase the efect by using multiple media. Figure 9 looks at
the levels of recommendation among diferent subgroups of people, categorised by their
exposure to specifc combinations of marketing. To make the analysis clear and focused, for
this particular chart we have looked only at current users of the brand, and considered
combinations of types of marketing. Word of mouth has been controlled to be equivalent for
each subgroup in the chart.
The results are extreme. Levels of recommendation range from only 13% among people who
have seen no marketing, through to 40% among people exposed to all three media. And
there is a subtlety in the fgures in between: there is no visible ‘tail-of’ in efect as we increase
the number of media, and the combination of digital and non-digital is more benefcial than
multiple types of traditional media.
Figure 9: Multiple media create the maximum efect
% recommending to others, within each group
All three
Other nondigital mktg AND
Digital mkg/website
TV AND digital mktg/website
TV AND other nondigital mktg
Digital mktg/website only
Other nondigital mktg only
TV only
None
0% 10% 20% 30% 40% 50%
Three
types
Two
types
One
type
None
Exposure to more media
increases likelihood to
recommend the brand
40%
34%
31%
27%
25%
24%
21%
13%
The analysis focuses on recommendation by brand users, for all
combinations of three types of media
12
The Infuence of Digital Communication...
This makes sense. Many advertisers deliberately reference the internet in their traditional
communications, inviting people to visit their website. Traditional media frst engage the
audience, then digital activity helps people along the pathway to purchase, or simply builds
on an initial level of interest. Prime examples of this efect would be campaigns like Cadbury’s
Gorilla or the Sony Bravia bouncing balls. The lesson from this analysis is to use the media in a
combination best suited to your brand and category, rather than put all your eggs into a
single basket.
Finally, again drawing from a diferent study that also used our touchpoint modelling, here is
a further reason why it is too early to proclaim the death of TV.
This is taken from a study we recently carried out in which our client had a campaign running
across a variety of media, from TV and Press creative, to online ads and Point of Sale material.
Everyone we spoke to in this study was in the market for the product – but when we look at
the infuences after running our touchpoint modelling, we can identify that those people
exposed to the TV work were much more emotionally impacted, while those exposed to the
online and point of sale material had bigger changes to their rational connections with the
brand.
Validation work from our EquiTrend study shows that for the vast majority of categories
it is important to build both an emotional and rational connection with the brand, in order to
attract and sustain a relationship with a customer. And in many categories, including
this example, it is necessary to stimulate the consumer to consider the whole idea of making
the purchase.
In this case, online advertising used on its own would have run the risk of reaching only a narrow
target audience, who were already somewhat pre-disposed to engage with the brand and seek
information. Traditional media used on its own, in particular TV, would have run the opposite
risk of awakening some interest but failing to follow through and ‘close the sale’.
Not shown in fgure 10, but identifed in the EquiTrend study was that word-of-mouth tends
to have a strong emotional reaction, understandably so given the tone of expression
employed in many word-of-mouth communications!
Figure 10: Choosing between media means not
choosing efect, but choosing between type of efect
TV
Press
Internet
POS
31%
40%
46%
59%
69%
60%
54%
41%
Rational Emotional
AHEAD OF WHAT’S NEXT.
The Infuence of Digital Communication...
Summary and Recommendations
In conclusion, let’s draw together these thoughts.
For all the rhetoric about online and social media, the ability of brands to dominate across a
medium with four billion pages is currently very limited; the reach of a brand’s campaign in
digital media will remain small, or fragmented, but its potential infuence among that group
could be large. Online therefore ofers the ‘audience of interest’ that traditional broadcast
fails to deliver efciently.
However, online works best when it is used to sharpen a message – a message that traditional
broadcast media are already carrying. While digital marketing ofers those users all the
information they need, traditional techniques ofer the emotional connection that is needed
to spark that need. Traditional media builds consideration: I’ve heard of that brand, and it is
worthy of my click.
In a world where social media is increasingly important, digital word-of-mouth is the epitome
of digital marketing: still very hard for a brand to achieve meaningful reach amidst the billions
of conversations, but intensely powerful when it does so. And that power needs to be
monitored, and indeed fed, (with content or comment) if brands are to properly harness it.
AHEAD OF WHAT’S NEXT.
© 2009 Harris Interactive, Inc. All Rights Reserved. 10.09
+44 (0) 20 8263 5200 [email protected] www.harrisinteractive.com/europe
doc_191314516.pdf