Description
The objective of this study is to explore the factors that are keys for an idea to be implemented on an
online crowdsourcing platform. A data set of 320 implemented ideas from My Starbucks Idea – an online
crowdsourcing platform – has been analyzed. We find that only one out of 500 users’ submitted ideas are
selected for implementation. The number of implemented ideas increases significantly at the early stage of
the platform. At the mature stage, even though an increasing number of ideas are submitted, implemented
ideas are proportionately low. Among the three categories of ideas – product, experience, and involvement
– ideas of the product category are implemented with lower values of some associated variables than that
of the experience category whereas those values in the involvement category are higher. Linked ideas
need lower scores than sole ideas to get implemented. The chance that an idea to be implemented largely
depends on votes received by and points earned on that idea
2214-4625/$ – see front matter © 2015 Holy Spirit University of Kaslik. Hosting by Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.aebj.2015.09.001
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111
Contents lists available at ScienceDirect
ScienceDirect
j our nal homepage: www. el sevi er. com/ l ocat e/ aebj
HOSTED BY
* Mokter Hossain. Tel.: +358451115611
E-mail address: [email protected]
Peer review under responsibility of Holy Spirit University of Kaslik.
Generating Ideas on Online Platforms: A Case Study of “My Starbucks
Idea”
Mokter Hossain
a
*, K. M. Zahidul Islam
b
a
Department of Industrial Engineering and Management, School of Science, Aalto University, Otaniementie 17, Espoo, 002150, Finland
b
Institute of Business Administration, Jahangirnagar University, Savar, Dhaka-1342, Bangladesh
AR T I CL E I NF O
Article history:
Received 16 April 15
Received in revised form 11 June 15
Accepted 12 June 15
Keywords:
Crowdsourcing
Idea generation
Idea implementation
Online platform
ABS T R AC T
The objective of this study is to explore the factors that are keys for an idea to be implemented on an
online crowdsourcing platform. A data set of 320 implemented ideas from My Starbucks Idea – an online
crowdsourcing platform – has been analyzed. We find that only one out of 500 users’ submitted ideas are
selected for implementation. The number of implemented ideas increases significantly at the early stage of
the platform. At the mature stage, even though an increasing number of ideas are submitted, implemented
ideas are proportionately low. Among the three categories of ideas – product, experience, and involvement
– ideas of the product category are implemented with lower values of some associated variables than that
of the experience category whereas those values in the involvement category are higher. Linked ideas
need lower scores than sole ideas to get implemented. The chance that an idea to be implemented largely
depends on votes received by and points earned on that idea.
© 2015 Holy Spirit University of Kaslik. Hosting by ElsevierB.V. All rights reserved.
1. Introduction
Business environments are increasingly becoming competitive. Consequently, firms are continuously seeking new ways of idea generation
(Westerski, Dalamagas & Iglesias, 2013). Researchers and practitioners are advocating for open approaches through engaging external individuals on
crowdsourcing platforms for innovation and ideation. Crowdsourcing is an option to get ideas from external sources. Howe (2006) coined the
crowdsourcing concept and defined it as follows: Crowdsourcing is the process of obtaining needed services, ideas, or contents by soliciting contributions
from a large group of people, and especially from an online community, rather than from traditional employees or suppliers. However, crowdsourcing has
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111 103
numerous definitions. For instance, Estellés-Arolas and González-Ladrón-de-Guevara (2012) found at least 40 definitions of crowdsourcing in the
literature.
The Internet serves as an important way for firms to collaborate with crowds for idea generation (Sawhney & Prandelli, 2000; Sawhney,
Verona & Prandelli, 2005). The importance of customers’ interaction for ideation is not a new phenomenon but the widespread availability of the Internet
has significantly increased the firms’ ability to interact with customers. Crowdsourcing has proven to be a great promise for ideation. Firms, irrespective
of small or large, are increasingly collaborating with external sources (Chesbrough, 2003; Dahlander & Wallin, 2006). Large firms such as Cisco, Dell,
Microsoft, Proctor and Gamble, Unilever and Starbucks are actively using crowdsourcing platforms to find ideas from external individuals such as users,
customers, amateurs, and volunteers (Di Gangi, Wasko & Hooker, 2010; Martínez-Torres, 2013; Westerski et al., 2013). On crowdsourcing platforms,
crowds not only interact with firms but also interact among themselves.
There are broadly two ways to solicit ideas from external sources on crowdsourcing platforms: (1) idea contest – calling external individuals to
submit ideas within a time period, selecting and awarding best ideas, and (2) ideation with continuous interactions between firms and crowds. This study
focuses on crowdsourcing where ideas are developed with continuous interactions between firms and crowds. In ideation with continuous interactions
between firms and crowds on online crowdsourcing platforms, crowds do not need to have high skills and expertise. Rather they can propose ideas based
on their day to day experiences. In other words, in the case of ideas which do not need significant cognition, active users can submit enormous number of
those types of ideas. For example, a user of Dell’s computer can submit ideas based on his/her using experience. Advanced information and
communication technologies (ICTs) along with Web 2.0 facilitate an environment for activities such as interaction, voting, comments, and discussion on
online crowdsourcing platforms (Hossain, 2012a; Mahr & Lievens, 2012). The ICTs with related tools and features have not only accelerated the quantity
but also the quality of ideas. Moreover, active users can be identified and enthused in ideation process to find more relevant ideas. Integrating customers
and other external individuals in the innovation process is considered as a powerful means to increase the success rate of and revenue from new offerings.
Yet, the understanding of the mechanism of customers’ integration for ideation is limitedly explored (Rohrbeck, Steinhoff & Perder, 2010).
Idea crowdsourcing possesses some challenges. Finding an appropriate team for idea evaluation is also a daunting task. Sometimes, some
excellent ideas may not be implementable by a firm (Martínez-Torres, 2013). High familiarity of an individual with a problem may block creativity and
identification of novel solutions (Franke, Poetz & Schreier, 2013; Wiley, 1998). Some managers resist crowdsourcing ideas because they are not sure what
kind of problem crowd can really solve and how to manage the whole process, and how to be sure that ideas received from crowds are the appropriate
ones (Boudreau & Lakhani, 2013). A common belief to engage professionals instead of crowd for idea generation is that they have the experience and
expertise to bring out promising ideas. In crowdsourcing, the idea evaluation process is very difficult especially when unimaginably large number of ideas
is generated in a short span of time (Jouret, 2009). The knowledge about crowds’ ideation skills is also sparse (Adamczyk, Bullinger & Möslein, 2012;
Bayus, 2013; Natalicchio, Petruzzelli & Garavelli, 2013).
Despite the importance of idea crowdsourcing, researchers have limited insights into this arena (Hossain & Kauranen, 2015; Poetz & Schreier,
2012). How the process of ideation is conducted effectively is a crucial issue for both researchers and practitioners (Schulze &Hoegl, 2008). To identify
best ideas from a large number of submitted ideas is a challenging issue for managers (Jouret, 2009). Moreover, how to identify and support the crowd
who are active in submitting implementable ideas is limitedly known (Kristensson & Magnusson, 2010). Hence, the objective of this study is to explore
the factors that are keys for an idea to be implemented through an online crowdsourcing platform.
2. Theoretical perspectives
The integration of customers as part of external innovation processes is crucial for firms (Enkel, Gassmann & Chesbrough, 2009). Poetz and
Schreier (2012) found that compared to in-house ideas, users’ product ideas place better positions in terms of novelty and customer benefits. They argue
that crowdsourcing might be considered as a promising method to gather external ideas which can, at least, complement internal idea generation of a firm.
In online crowdsourcing platforms with continuous interactions between firms and crowds, the selection of an idea may depend on various factors, such as
votes, comments, point earned, the amount of submitted ideas, relevance, feasibility to implement, and alignment of an idea with a firm’s business
strategy. A platform management team directs the overall ideation process so that an idea is refined, highly voted, and widely accepted by platform
members before its implementation. Great ideas may get immediate attention from both community members and selection teams. Consequently, these
ideas may become central issues on a platform. Moreover, an ideator’s contribution to others’ ideas such as voting, commenting, discussing, etc. helps to
get attention to his/her own idea from other community members as well as from platform management team.
In general, online crowdsourcing platforms of large firms receive numerous ideas. However, the number of implementable ideas is very
limited. Hence, most of those ideas remain unused. In many cases, huge amounts of information submitted by crowd have negative effect in the
identification of most promising ideas and a poorly managed online crowdsourcing platform may result in a fiasco (Di Gangi, Wasko & Hooker, 2010).
Prior studies pointed out that if a firm is not able to attract right crowds who could provide valuable ideas, the successful idea generation is seemingly
impossible (Piller & Walcher, 2006). Generally, a small number of the crowds provide valuable ideas, and majority of the crowds are involved with
activities such as making comments, casting votes and providing suggestions. It is important to identify individuals who are very active and enthusiast to
104 ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111
submit implementable ideas (Martínez-Torres, 2012). On platforms, users express their experiences, raise questions, comments, vote on proposed ideas,
and answer questions posed by others. Thus, they develop ideas as a community (Rowley, Kupiec-Teahan & Leeming, 2007).
Idea generation, in many cases, goes far beyond the imagination of crowds. Some scholars believe that customers may remain too absorbed in
the existing products that may prevent them from coming up with truly novel ideas (see Leonard & Rayport, 1997). However, collaboration with
customers helps firms to create value through product innovation (Sawhney, Verona & Prandelli, 2005). Firms can tap into customer knowledge and
engage them for ideation through continuous interactions (Nambisan, 2002; Sawhney & Prandelli, 2000). Assessing ideas to select appropriate ideas is a
challenging task. Hrastinski, Kviselius, Ozan and Edenius (2010) point out that ideas from a large number of submitted ideas on platforms are typically
selected in several ways: using some very simple community statistics, expert reviews of the submitted ideas and the number of votes or comments an idea
received. A large number of ideas from crowd raise the challenge of absorptive capacity – firms’ ability to identify and evaluate new knowledge into the
current business (see Cohen &Levinthal, 1990). On the other hand, Stevens and Burley (1997) propound a success curve and claimed that it is valid for
most industries regarding idea success. They argue that only one out of 3000 ideas ultimately becomes commercially successful.
The Motivation of a crowd to participate in online platform is a key factor for the success of a platform. Intrinsic motivation is more prevalent
than extrinsic motivation when the task of individuals is simple. Online platforms of many large firms are based primarily on intrinsic motivation. Apple,
for example, has turned towards crowd to propel its growth. Starbucks’ My Starbucks Idea, IBM’s Global Innovation Jam, and Dell’s IdeaStorm are
highly popular online crowdsourcing platforms which are based on intrinsic motivation. Intermediary platforms are also operating with various business
models (see Hossain, 2012a; Hossain and Kauranen, 2014). In some cases, some kind of incentive prize is offered to individuals whose ideas are
implemented. Intrinsic motivation includes personal learning, expression, creativity, enjoyment, fun, entertainment, and care of a community, among
others (Antikainen & Vaataja, 2010; Boudreau & Lakhani, 2009; Hossain, 2012b).
3. Materials and methodology
3.1. Description of the case platform
My Starbucks Idea is considered as the case for this study. It was launched in March, 2008. Crowd can submit their ideas, vote and make
comments on submitted ideas of others, etc. It is a place of interaction between Starbucks and crowds to improve the organization as a whole. This
platform provides customers opportunities to express their views to improve offerings. Anyone can register to join in the crowdsourcing platform with
valid credential at free of cost. As on August 20, 2013, customers have submitted 162,156 ideas (see Table 1). Submitted ideas are classified into three
categories: product, experience, and involvement. On the platform, crowds discuss, debate, and argue on various topics related with Starbucks’ products
and services.
Platform management team provides comments to lead discussion and other activities to the right direction so that the crowd can submit more
implementable ideas. Most of the ideas are under the category of product (105,161) followed by experience (35,098) while the number of ideas under
involvement category is the lowest (21,897). Among 162,156 submitted ideas, only 320 ideas have been implemented. In other words, one out of around
500 ideas finds its way to Starbucks store after passing through the crowd and the firm’s evaluation process. Of the 320 implemented ideas, 255 ideas
belong to product category, 46 to experience category, and 19 to involvement category.
Table 1- List of ideas submitted by crowd.
Source:http://mystarbucksidea.force.com/ (August 20, 2013).
Under the frequently asked questions (FAQ) part of the platform, Starbucks has mentioned clearly various regulations of participation into the
community. Any submitted idea becomes a property of Starbucks and no compensation is promised. Idea submission is voluntary, non-confidential, non-
Product Ideas (#) Involvement Ideas (#) Experience Ideas (#)
Coffee & Espresso Drinks 34,542 Other Involvement Ideas 5,686 Atmosphere & Locations 15,294
Starbucks Card 17,063 Building Community 5,215 Other Experience Ideas 11,487
Food 16,267 Outside USA 1,626 Ordering, Payment, & Pick-Up 8,317
Other Product Ideas 11,202 Social Responsibility 9,37
Tea & Other Drinks 10,196
Merchandise & Music 8,464
Frappuccino® Beverages 4,066
New Technology 3,361
Total Ideas in each category 105,161 21,897 35,098
Total Ideas in three categories 162,156
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111 105
committal, gratuitous, perpetual, irrevocable and non-exclusive. Starbucks gets royalty-free license to use any ideas or other contribution. By far,
Starbucks has become very successful to generate promising ideas. Its Facebook account has over 35 million fans. Any posted document on the platform
is shared to all other members as long as the content does not breach the rules and regulations of the platform. Starbucks does not share privately held
information. However, some activities of the platform are highly criticized. For example, ideas are not properly catalogued and reviewed; hence a previous
idea may resurface after a long period of time (Rosen, 2011).
3.2. Data collection
The implemented ideas are separately listed on the platform. The necessary and possible information related with each implemented idea are
extracted manually from the Website by visiting each idea link. Extracted information related with each implemented idea are: vote received, point earned
by idea submitter, point earned on an idea, comments received, category of idea (product, experience, and involvement), sole idea or an idea related with
other submitted ideas. The extracted information is recorded in a spreadsheet for analysis purpose. Additionally, the registration dates of crowds of the
implemented ideas and the dates of implementation of their ideas are also recorded. Moreover, names of the ideas with related statement are recorded. To
find the duration of a crowd’s involvement with the platform, we subtracted the date of registration of that crowd, whose ideas were implemented, from
the date August 1, 2013. This later date is assumed as a benchmark to understand the relative period of a user’s involvement with the platform.
3.3. Variables construction
We extracted variables based on data collected from the crowdsourcing platform. Ideas are classified into three major categories: product,
experience, and involvement, which are considered as three different dependent variables to be used in three models. Moreover, some ideas are linked
with other already submitted ideas. Based on link of ideas, they are classified into two groups: linked ideas and sole ideas. We used both linked ideas and
sole ideas as two dependent variables to compare the ideas of these two groups. The platform management team finally decides to implement an idea after
assessing an idea’s potential in terms of the company’s strategy and customers’ benefits. However, an idea is implemented based on some quantitative
values which are used as dependent variables here. Table 2 outlines the description of the variables used in the study.
Table 2- Description of variables.
Variables Descriptions
Dependent variable
Product Category of ideas that are product type
Experience Category of ideas that are experience type
Involvement Category of ideas that are involvement type
Linked ideas Group of ideas which are linked with other ideas
Sole ideas Group of ideas which are not linked with other ideas
Independent variable
User period of presence Period from the time of registration of a crowd to a baseline
Ideas submitted Number of ideas submitted by a particular ideator
Submission to implementation Period between from the date of idea submission to its date of implementation
Vote submitted by an ideator Number of votes by an ideator
Vote received on ideas Number of vote received on ideas
Comments submitted by an ideator Number of comments submitted by an ideator on the ideas of others
Point earned Number of points earned by an ideator
Comment received Number of comments received on ideas
Point on idea Number of points received on idea
106 ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111
We used the period of presence of an ideator as an independent variable to see how the period of presence of an ideator affects in generating
implementable ideas. An idea may get implemented instantly if it is considered as a great idea and easy to implement. Thus, the period from the
submission to the implementation of an idea is an important variable that affects an idea to be implemented. An ideator’s performance in the eyes of peers
on a crowdsourcing platform is indicated by the number of points he or she earned from other crowds. Factors such as the number of ideas, the number of
votes, and the number of comments submitted by a particular ideator play a crucial role to assess the activeness of an ideator. On the other hand, an idea is
preliminary measured based on some factors such as the number of votes received, the number of points earned, the number of comments received, and
the points earned on that idea. Hence, we considered these factors as independent variables. Thus, we have considered nine independent variables to
understand their influence on an idea to be implemented.
3.4. Data analysis
We used both parametric (t-test and ANOVA test) and non-parametric tests (Mann-Whitney U and Kruskal-Wallis tests) to look for the
robustness of the estimated results. The independent sample t-test was used to test whether the two groups of ideas – linked and sole ideas – are
independent of each other in the obvious sense that they are separate samples containing different sets of individual characteristics. We did ANOVA test
to see if multiple means of different variables in three categories of ideas are equal to each other. On the other hand, the non-parametric Mann-Whitney U
test and the Kruskal-Wallis test were applied to detect whether two or more categories of samples come from the same distribution based on median
values under the assumption that the shapes of the underlying distributions are the same. Finally, we applied a multiple regression analysis to identify the
determinants of an idea’s implementation.
4. Results and analysis
Figure 1 illustrates the number of implemented ideas over a five-year period. Since the platform was launched in March 2008, the figure of
implemented ideas for 2008 is of eight months (April to December). The number of implemented ideas remained almost the same in the first two years. In
2010, there was a significant jump in the number of idea implementations and it had increased steadily in the subsequent years.
Fig. 1- Trend of idea implementation over years based on Starbucks’ record.
Table 3 presents the results of descriptive statistics and correlations between the considered variables. Highly significant positive correlations
exist between variables, such as comments submitted and votes submitted; points earned and comments submitted; and points earned and votes submitted.
Moreover, variables such as comments received and points of idea, ideas submitted and points earned, vote received and idea submitted are also
significantly positively correlated. The period of presence of users has significant positive correlation with variables such as submitted to implement, votes
received, comments received, and points on idea. Additionally, the variable vote received is positively related with variables, such as comments submitted,
comments received, and points on idea at a significant level. Overall, the correlation matrix shows that there is a considerable number of variables which
are highly correlated with each other as we see in Table 3.
0
10
20
30
40
50
60
70
80
2008 2009 2010 2011 2012
I
m
p
l
e
m
e
n
e
t
e
d
I
d
e
a
s
Year
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111 107
Table 3- Descriptive statistics and simple correlation.
Variables Mean SD 1 2 3 4 5 6 7 8 9
1 User Period of presence 1516.58 6.3749 1
2 Ideas submitted 4.40 11.74 0.08 1
3 Submitted to implement 395.18
378.28
0.29**
0.03
1
4 Vote submitted 19.32 165.45 -0.10 0.36** -0.02 1
5 Vote received
1624.21
5548.41
0.14**
0.57**
-0.01
0.06
1
6 Comments submitted
2.87
17.30
-0.06
0.53**
-0.02
0.95**
0.18** 1
7 Point earned
38.19
204.54
-0.04
0.60**
-0.01
0.93**
0.37**
0.94**
1
8 Comment received
25.28
66.72
0.15**
0.04
0.03
-0.09
0.22**
0.07
0.05
1
9 Point on idea 5092.89
12432.43
0.17**
0.01
0.06
-0.02
0.3**
-0.02
0.07
0.61**
1
Notes: **P
The objective of this study is to explore the factors that are keys for an idea to be implemented on an
online crowdsourcing platform. A data set of 320 implemented ideas from My Starbucks Idea – an online
crowdsourcing platform – has been analyzed. We find that only one out of 500 users’ submitted ideas are
selected for implementation. The number of implemented ideas increases significantly at the early stage of
the platform. At the mature stage, even though an increasing number of ideas are submitted, implemented
ideas are proportionately low. Among the three categories of ideas – product, experience, and involvement
– ideas of the product category are implemented with lower values of some associated variables than that
of the experience category whereas those values in the involvement category are higher. Linked ideas
need lower scores than sole ideas to get implemented. The chance that an idea to be implemented largely
depends on votes received by and points earned on that idea
2214-4625/$ – see front matter © 2015 Holy Spirit University of Kaslik. Hosting by Elsevier B.V. All rights reserved.http://dx.doi.org/10.1016/j.aebj.2015.09.001
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111
Contents lists available at ScienceDirect
ScienceDirect
j our nal homepage: www. el sevi er. com/ l ocat e/ aebj
HOSTED BY
* Mokter Hossain. Tel.: +358451115611
E-mail address: [email protected]
Peer review under responsibility of Holy Spirit University of Kaslik.
Generating Ideas on Online Platforms: A Case Study of “My Starbucks
Idea”
Mokter Hossain
a
*, K. M. Zahidul Islam
b
a
Department of Industrial Engineering and Management, School of Science, Aalto University, Otaniementie 17, Espoo, 002150, Finland
b
Institute of Business Administration, Jahangirnagar University, Savar, Dhaka-1342, Bangladesh
AR T I CL E I NF O
Article history:
Received 16 April 15
Received in revised form 11 June 15
Accepted 12 June 15
Keywords:
Crowdsourcing
Idea generation
Idea implementation
Online platform
ABS T R AC T
The objective of this study is to explore the factors that are keys for an idea to be implemented on an
online crowdsourcing platform. A data set of 320 implemented ideas from My Starbucks Idea – an online
crowdsourcing platform – has been analyzed. We find that only one out of 500 users’ submitted ideas are
selected for implementation. The number of implemented ideas increases significantly at the early stage of
the platform. At the mature stage, even though an increasing number of ideas are submitted, implemented
ideas are proportionately low. Among the three categories of ideas – product, experience, and involvement
– ideas of the product category are implemented with lower values of some associated variables than that
of the experience category whereas those values in the involvement category are higher. Linked ideas
need lower scores than sole ideas to get implemented. The chance that an idea to be implemented largely
depends on votes received by and points earned on that idea.
© 2015 Holy Spirit University of Kaslik. Hosting by ElsevierB.V. All rights reserved.
1. Introduction
Business environments are increasingly becoming competitive. Consequently, firms are continuously seeking new ways of idea generation
(Westerski, Dalamagas & Iglesias, 2013). Researchers and practitioners are advocating for open approaches through engaging external individuals on
crowdsourcing platforms for innovation and ideation. Crowdsourcing is an option to get ideas from external sources. Howe (2006) coined the
crowdsourcing concept and defined it as follows: Crowdsourcing is the process of obtaining needed services, ideas, or contents by soliciting contributions
from a large group of people, and especially from an online community, rather than from traditional employees or suppliers. However, crowdsourcing has
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111 103
numerous definitions. For instance, Estellés-Arolas and González-Ladrón-de-Guevara (2012) found at least 40 definitions of crowdsourcing in the
literature.
The Internet serves as an important way for firms to collaborate with crowds for idea generation (Sawhney & Prandelli, 2000; Sawhney,
Verona & Prandelli, 2005). The importance of customers’ interaction for ideation is not a new phenomenon but the widespread availability of the Internet
has significantly increased the firms’ ability to interact with customers. Crowdsourcing has proven to be a great promise for ideation. Firms, irrespective
of small or large, are increasingly collaborating with external sources (Chesbrough, 2003; Dahlander & Wallin, 2006). Large firms such as Cisco, Dell,
Microsoft, Proctor and Gamble, Unilever and Starbucks are actively using crowdsourcing platforms to find ideas from external individuals such as users,
customers, amateurs, and volunteers (Di Gangi, Wasko & Hooker, 2010; Martínez-Torres, 2013; Westerski et al., 2013). On crowdsourcing platforms,
crowds not only interact with firms but also interact among themselves.
There are broadly two ways to solicit ideas from external sources on crowdsourcing platforms: (1) idea contest – calling external individuals to
submit ideas within a time period, selecting and awarding best ideas, and (2) ideation with continuous interactions between firms and crowds. This study
focuses on crowdsourcing where ideas are developed with continuous interactions between firms and crowds. In ideation with continuous interactions
between firms and crowds on online crowdsourcing platforms, crowds do not need to have high skills and expertise. Rather they can propose ideas based
on their day to day experiences. In other words, in the case of ideas which do not need significant cognition, active users can submit enormous number of
those types of ideas. For example, a user of Dell’s computer can submit ideas based on his/her using experience. Advanced information and
communication technologies (ICTs) along with Web 2.0 facilitate an environment for activities such as interaction, voting, comments, and discussion on
online crowdsourcing platforms (Hossain, 2012a; Mahr & Lievens, 2012). The ICTs with related tools and features have not only accelerated the quantity
but also the quality of ideas. Moreover, active users can be identified and enthused in ideation process to find more relevant ideas. Integrating customers
and other external individuals in the innovation process is considered as a powerful means to increase the success rate of and revenue from new offerings.
Yet, the understanding of the mechanism of customers’ integration for ideation is limitedly explored (Rohrbeck, Steinhoff & Perder, 2010).
Idea crowdsourcing possesses some challenges. Finding an appropriate team for idea evaluation is also a daunting task. Sometimes, some
excellent ideas may not be implementable by a firm (Martínez-Torres, 2013). High familiarity of an individual with a problem may block creativity and
identification of novel solutions (Franke, Poetz & Schreier, 2013; Wiley, 1998). Some managers resist crowdsourcing ideas because they are not sure what
kind of problem crowd can really solve and how to manage the whole process, and how to be sure that ideas received from crowds are the appropriate
ones (Boudreau & Lakhani, 2013). A common belief to engage professionals instead of crowd for idea generation is that they have the experience and
expertise to bring out promising ideas. In crowdsourcing, the idea evaluation process is very difficult especially when unimaginably large number of ideas
is generated in a short span of time (Jouret, 2009). The knowledge about crowds’ ideation skills is also sparse (Adamczyk, Bullinger & Möslein, 2012;
Bayus, 2013; Natalicchio, Petruzzelli & Garavelli, 2013).
Despite the importance of idea crowdsourcing, researchers have limited insights into this arena (Hossain & Kauranen, 2015; Poetz & Schreier,
2012). How the process of ideation is conducted effectively is a crucial issue for both researchers and practitioners (Schulze &Hoegl, 2008). To identify
best ideas from a large number of submitted ideas is a challenging issue for managers (Jouret, 2009). Moreover, how to identify and support the crowd
who are active in submitting implementable ideas is limitedly known (Kristensson & Magnusson, 2010). Hence, the objective of this study is to explore
the factors that are keys for an idea to be implemented through an online crowdsourcing platform.
2. Theoretical perspectives
The integration of customers as part of external innovation processes is crucial for firms (Enkel, Gassmann & Chesbrough, 2009). Poetz and
Schreier (2012) found that compared to in-house ideas, users’ product ideas place better positions in terms of novelty and customer benefits. They argue
that crowdsourcing might be considered as a promising method to gather external ideas which can, at least, complement internal idea generation of a firm.
In online crowdsourcing platforms with continuous interactions between firms and crowds, the selection of an idea may depend on various factors, such as
votes, comments, point earned, the amount of submitted ideas, relevance, feasibility to implement, and alignment of an idea with a firm’s business
strategy. A platform management team directs the overall ideation process so that an idea is refined, highly voted, and widely accepted by platform
members before its implementation. Great ideas may get immediate attention from both community members and selection teams. Consequently, these
ideas may become central issues on a platform. Moreover, an ideator’s contribution to others’ ideas such as voting, commenting, discussing, etc. helps to
get attention to his/her own idea from other community members as well as from platform management team.
In general, online crowdsourcing platforms of large firms receive numerous ideas. However, the number of implementable ideas is very
limited. Hence, most of those ideas remain unused. In many cases, huge amounts of information submitted by crowd have negative effect in the
identification of most promising ideas and a poorly managed online crowdsourcing platform may result in a fiasco (Di Gangi, Wasko & Hooker, 2010).
Prior studies pointed out that if a firm is not able to attract right crowds who could provide valuable ideas, the successful idea generation is seemingly
impossible (Piller & Walcher, 2006). Generally, a small number of the crowds provide valuable ideas, and majority of the crowds are involved with
activities such as making comments, casting votes and providing suggestions. It is important to identify individuals who are very active and enthusiast to
104 ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111
submit implementable ideas (Martínez-Torres, 2012). On platforms, users express their experiences, raise questions, comments, vote on proposed ideas,
and answer questions posed by others. Thus, they develop ideas as a community (Rowley, Kupiec-Teahan & Leeming, 2007).
Idea generation, in many cases, goes far beyond the imagination of crowds. Some scholars believe that customers may remain too absorbed in
the existing products that may prevent them from coming up with truly novel ideas (see Leonard & Rayport, 1997). However, collaboration with
customers helps firms to create value through product innovation (Sawhney, Verona & Prandelli, 2005). Firms can tap into customer knowledge and
engage them for ideation through continuous interactions (Nambisan, 2002; Sawhney & Prandelli, 2000). Assessing ideas to select appropriate ideas is a
challenging task. Hrastinski, Kviselius, Ozan and Edenius (2010) point out that ideas from a large number of submitted ideas on platforms are typically
selected in several ways: using some very simple community statistics, expert reviews of the submitted ideas and the number of votes or comments an idea
received. A large number of ideas from crowd raise the challenge of absorptive capacity – firms’ ability to identify and evaluate new knowledge into the
current business (see Cohen &Levinthal, 1990). On the other hand, Stevens and Burley (1997) propound a success curve and claimed that it is valid for
most industries regarding idea success. They argue that only one out of 3000 ideas ultimately becomes commercially successful.
The Motivation of a crowd to participate in online platform is a key factor for the success of a platform. Intrinsic motivation is more prevalent
than extrinsic motivation when the task of individuals is simple. Online platforms of many large firms are based primarily on intrinsic motivation. Apple,
for example, has turned towards crowd to propel its growth. Starbucks’ My Starbucks Idea, IBM’s Global Innovation Jam, and Dell’s IdeaStorm are
highly popular online crowdsourcing platforms which are based on intrinsic motivation. Intermediary platforms are also operating with various business
models (see Hossain, 2012a; Hossain and Kauranen, 2014). In some cases, some kind of incentive prize is offered to individuals whose ideas are
implemented. Intrinsic motivation includes personal learning, expression, creativity, enjoyment, fun, entertainment, and care of a community, among
others (Antikainen & Vaataja, 2010; Boudreau & Lakhani, 2009; Hossain, 2012b).
3. Materials and methodology
3.1. Description of the case platform
My Starbucks Idea is considered as the case for this study. It was launched in March, 2008. Crowd can submit their ideas, vote and make
comments on submitted ideas of others, etc. It is a place of interaction between Starbucks and crowds to improve the organization as a whole. This
platform provides customers opportunities to express their views to improve offerings. Anyone can register to join in the crowdsourcing platform with
valid credential at free of cost. As on August 20, 2013, customers have submitted 162,156 ideas (see Table 1). Submitted ideas are classified into three
categories: product, experience, and involvement. On the platform, crowds discuss, debate, and argue on various topics related with Starbucks’ products
and services.
Platform management team provides comments to lead discussion and other activities to the right direction so that the crowd can submit more
implementable ideas. Most of the ideas are under the category of product (105,161) followed by experience (35,098) while the number of ideas under
involvement category is the lowest (21,897). Among 162,156 submitted ideas, only 320 ideas have been implemented. In other words, one out of around
500 ideas finds its way to Starbucks store after passing through the crowd and the firm’s evaluation process. Of the 320 implemented ideas, 255 ideas
belong to product category, 46 to experience category, and 19 to involvement category.
Table 1- List of ideas submitted by crowd.
Source:http://mystarbucksidea.force.com/ (August 20, 2013).
Under the frequently asked questions (FAQ) part of the platform, Starbucks has mentioned clearly various regulations of participation into the
community. Any submitted idea becomes a property of Starbucks and no compensation is promised. Idea submission is voluntary, non-confidential, non-
Product Ideas (#) Involvement Ideas (#) Experience Ideas (#)
Coffee & Espresso Drinks 34,542 Other Involvement Ideas 5,686 Atmosphere & Locations 15,294
Starbucks Card 17,063 Building Community 5,215 Other Experience Ideas 11,487
Food 16,267 Outside USA 1,626 Ordering, Payment, & Pick-Up 8,317
Other Product Ideas 11,202 Social Responsibility 9,37
Tea & Other Drinks 10,196
Merchandise & Music 8,464
Frappuccino® Beverages 4,066
New Technology 3,361
Total Ideas in each category 105,161 21,897 35,098
Total Ideas in three categories 162,156
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111 105
committal, gratuitous, perpetual, irrevocable and non-exclusive. Starbucks gets royalty-free license to use any ideas or other contribution. By far,
Starbucks has become very successful to generate promising ideas. Its Facebook account has over 35 million fans. Any posted document on the platform
is shared to all other members as long as the content does not breach the rules and regulations of the platform. Starbucks does not share privately held
information. However, some activities of the platform are highly criticized. For example, ideas are not properly catalogued and reviewed; hence a previous
idea may resurface after a long period of time (Rosen, 2011).
3.2. Data collection
The implemented ideas are separately listed on the platform. The necessary and possible information related with each implemented idea are
extracted manually from the Website by visiting each idea link. Extracted information related with each implemented idea are: vote received, point earned
by idea submitter, point earned on an idea, comments received, category of idea (product, experience, and involvement), sole idea or an idea related with
other submitted ideas. The extracted information is recorded in a spreadsheet for analysis purpose. Additionally, the registration dates of crowds of the
implemented ideas and the dates of implementation of their ideas are also recorded. Moreover, names of the ideas with related statement are recorded. To
find the duration of a crowd’s involvement with the platform, we subtracted the date of registration of that crowd, whose ideas were implemented, from
the date August 1, 2013. This later date is assumed as a benchmark to understand the relative period of a user’s involvement with the platform.
3.3. Variables construction
We extracted variables based on data collected from the crowdsourcing platform. Ideas are classified into three major categories: product,
experience, and involvement, which are considered as three different dependent variables to be used in three models. Moreover, some ideas are linked
with other already submitted ideas. Based on link of ideas, they are classified into two groups: linked ideas and sole ideas. We used both linked ideas and
sole ideas as two dependent variables to compare the ideas of these two groups. The platform management team finally decides to implement an idea after
assessing an idea’s potential in terms of the company’s strategy and customers’ benefits. However, an idea is implemented based on some quantitative
values which are used as dependent variables here. Table 2 outlines the description of the variables used in the study.
Table 2- Description of variables.
Variables Descriptions
Dependent variable
Product Category of ideas that are product type
Experience Category of ideas that are experience type
Involvement Category of ideas that are involvement type
Linked ideas Group of ideas which are linked with other ideas
Sole ideas Group of ideas which are not linked with other ideas
Independent variable
User period of presence Period from the time of registration of a crowd to a baseline
Ideas submitted Number of ideas submitted by a particular ideator
Submission to implementation Period between from the date of idea submission to its date of implementation
Vote submitted by an ideator Number of votes by an ideator
Vote received on ideas Number of vote received on ideas
Comments submitted by an ideator Number of comments submitted by an ideator on the ideas of others
Point earned Number of points earned by an ideator
Comment received Number of comments received on ideas
Point on idea Number of points received on idea
106 ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111
We used the period of presence of an ideator as an independent variable to see how the period of presence of an ideator affects in generating
implementable ideas. An idea may get implemented instantly if it is considered as a great idea and easy to implement. Thus, the period from the
submission to the implementation of an idea is an important variable that affects an idea to be implemented. An ideator’s performance in the eyes of peers
on a crowdsourcing platform is indicated by the number of points he or she earned from other crowds. Factors such as the number of ideas, the number of
votes, and the number of comments submitted by a particular ideator play a crucial role to assess the activeness of an ideator. On the other hand, an idea is
preliminary measured based on some factors such as the number of votes received, the number of points earned, the number of comments received, and
the points earned on that idea. Hence, we considered these factors as independent variables. Thus, we have considered nine independent variables to
understand their influence on an idea to be implemented.
3.4. Data analysis
We used both parametric (t-test and ANOVA test) and non-parametric tests (Mann-Whitney U and Kruskal-Wallis tests) to look for the
robustness of the estimated results. The independent sample t-test was used to test whether the two groups of ideas – linked and sole ideas – are
independent of each other in the obvious sense that they are separate samples containing different sets of individual characteristics. We did ANOVA test
to see if multiple means of different variables in three categories of ideas are equal to each other. On the other hand, the non-parametric Mann-Whitney U
test and the Kruskal-Wallis test were applied to detect whether two or more categories of samples come from the same distribution based on median
values under the assumption that the shapes of the underlying distributions are the same. Finally, we applied a multiple regression analysis to identify the
determinants of an idea’s implementation.
4. Results and analysis
Figure 1 illustrates the number of implemented ideas over a five-year period. Since the platform was launched in March 2008, the figure of
implemented ideas for 2008 is of eight months (April to December). The number of implemented ideas remained almost the same in the first two years. In
2010, there was a significant jump in the number of idea implementations and it had increased steadily in the subsequent years.
Fig. 1- Trend of idea implementation over years based on Starbucks’ record.
Table 3 presents the results of descriptive statistics and correlations between the considered variables. Highly significant positive correlations
exist between variables, such as comments submitted and votes submitted; points earned and comments submitted; and points earned and votes submitted.
Moreover, variables such as comments received and points of idea, ideas submitted and points earned, vote received and idea submitted are also
significantly positively correlated. The period of presence of users has significant positive correlation with variables such as submitted to implement, votes
received, comments received, and points on idea. Additionally, the variable vote received is positively related with variables, such as comments submitted,
comments received, and points on idea at a significant level. Overall, the correlation matrix shows that there is a considerable number of variables which
are highly correlated with each other as we see in Table 3.
0
10
20
30
40
50
60
70
80
2008 2009 2010 2011 2012
I
m
p
l
e
m
e
n
e
t
e
d
I
d
e
a
s
Year
ARAB ECONOMICS AND BUSINESS JOURNAL 10 (2015) 102–111 107
Table 3- Descriptive statistics and simple correlation.
Variables Mean SD 1 2 3 4 5 6 7 8 9
1 User Period of presence 1516.58 6.3749 1
2 Ideas submitted 4.40 11.74 0.08 1
3 Submitted to implement 395.18
378.28
0.29**
0.03
1
4 Vote submitted 19.32 165.45 -0.10 0.36** -0.02 1
5 Vote received
1624.21
5548.41
0.14**
0.57**
-0.01
0.06
1
6 Comments submitted
2.87
17.30
-0.06
0.53**
-0.02
0.95**
0.18** 1
7 Point earned
38.19
204.54
-0.04
0.60**
-0.01
0.93**
0.37**
0.94**
1
8 Comment received
25.28
66.72
0.15**
0.04
0.03
-0.09
0.22**
0.07
0.05
1
9 Point on idea 5092.89
12432.43
0.17**
0.01
0.06
-0.02
0.3**
-0.02
0.07
0.61**
1
Notes: **P