For IKEA, the yearly Catalogue is the main communication channel with existing and potential customers globally. This case study shows how the 2013 edition of the Catalogue and possible covers for the 2014 edition were evaluated qualitatively around the world, through Market Research Online Communities (or Consumer Consulting Boards) in five different countries.
2. The paper also shows how ‘Online Communities’ are becoming a
true ‘Fusion Research’ tool more and more which allows for
‘triangulation’on different levels (data sources, research methods,
research environments, theories and investigators),
leading to more valid research results, fresh
inspiration and a deeper understanding of the
issue researched. Best practices concerning:
moving an existing qualitative project online,
creating internal buy-in for emerging research
methods, engaging internal audiences with research
findings, running communities in different cultures
and reactivating an MROC over time, are shared
as well.
What to
expect?
For IKEA, the yearly Catalogue is the main communication channel
with existing and potential customers globally. This case study show
how the 2013 edition of the Catalogue and possible covers for the
2014 edition were evaluated qualitatively around the world,
through ‘Market Research Online Communities’ (or ‘Consumer Consulting
Boards’) in five different countries.
The IKEA Catalogue
4. IKEA has the vision ‘to create a better
everyday life for many people’ by ‘offering
a wide range of well-designed, functional
home furnishing products at prices so low
that as many people as possible will be able
to afford them’. For IKEA, their yearly
Catalogue is one of the main channels of
communication with existing and potential
customers globally: one Catalogue showcasing
one product offering to serve ‘the many
people’ around the globe. Every year, IKEA
conducts a ‘global’ qualitative study to
understand how people feel and think about the
latest edition of the Catalogue and what they do
with it (this was done before the actual launch).
A few weeks after the global distribution, a large quantitative
study was conducted to evaluate the new edition on key
performance indicators, to measure the impact it had in
the market and to benchmark the performance with
previous editions. The findings of these studies serve as input
for the creative team that works on the latest edition.
For the 2013 edition, significant changes were made to
the format (slightly bigger), content (offering more inspiration
pages, including more storytelling and a different way of
picturing the products) and structure of the Catalogue (Figure
1). Furthermore, a complementary mobile application was
launched. These structural changes made it even more
important to deeply understand people’s emotional and
rational reactions to the latest edition.
6. 3
2
1
Over the past years the feeling had grown within IKEA that the qualitative evaluation of the Catalogue, done
through offline focus groups in different countries around the world, could potentially be done differently
and better and in a fresher way. Here is a shortlist of the key issues that the business owners and the research team
at IKEA were encountering:
Focus groups were only giving a snapshot of reality: the first reactions to the new Catalogue. No real
insights were gained on people’s second thoughts and their behaviour afterwards (how they are using the
Catalogue and how that use evolved over time).
Secondly, given the limited time spent with consumers and a limited portfolio of research techniques possible,
focus groups did not bring that much fresh and inspiring information to the table. Every year, a new
edition of the Catalogue was only judged partially (due to the limits of the method) and in the same way,
leading to conclusions that were very similar year after year.
Another issue flagged by internal stakeholders was the limited number of participants per session in a
focus group and the dominance of certain individuals in the discussion. Communities give access
to more opinions of people with a wide range of profiles. And due to the longer time period, all participants
are given an equal chance to give their opinion.
7. 5
4 Fourthly, the project owners experienced that the quality of a focus group was largely
determined by the quality and experience of the moderator. There was probably interviewer bias
as well. Country differences were possibly due to differences in moderation style, rather than real
differences between cultures.
Finally, despite all the new technologies available to follow offline focus groups from home as a client,
internal stakeholders do not follow sessions that often.
Focus group transcripts are experienced as not that
convenient and pleasant to go back to the real discussion for
inspiration or to justify an element when creating or
making decisions. The result is that internal stakeholders
are less confronted with the ‘voice of the customers’ during
and after the project.
Conclusion:
8. Given this situation, the research team at IKEA was on
the look-out for a method enabling them to
understand both the emotional and rational
reactions to the 2013 Catalogue and the mobile
application. Furthermore, in order to really understand
the Catalogue’s impact, it was also important to gain
insights into how people use the Catalogue on a daily
basis after it landed in their homes.
Given these objectives, a Market Research Online
Community (MROC) (De Ruyck et al, 2010) was chosen
as the backbone of this project in favour of traditional
focus groups (See Figure 2). Online closed platforms
to have a dialogue and work together with
consumers in five different countries around the
world (Germany, Italy, Poland, the US and China)
seemed to be the answer to the above issues. But why
was this really a better option?
A simple, but very useful framework to evaluate a new research method
and to demonstrate its effectiveness internally is to check if it is
providing ‘automatical’, ‘informational’ and/or
‘transformational’ benefits
9. 1 Automational: doing things faster and more cost-efficient
A research community indeed takes more time to set
up, but once created it can be reused without losing
the time of a classical set-up phase. This is a plus in
the process of crafting a new Catalogue: the platform
cannot only be used to evaluate the current issue, it
can also be used as a source of feedback and
inspiration during the process of creating the next
edition. For this project five communities with a 3-
week duration were created. Eight weeks later the
platforms were reopened for one week to get feedback
on the next stage - the creation of the cover for the
2014 edition.
Secondly, the set-up cost of a community project is
higher than that of an ad hoc qualitative project, but once
established and in use it becomes a cost-efficient tool:
more and different research methods can be
combined within the same budget.
The @Home Community
10. 2 Informational: obtaining a better data quality and deeper insights
Communities are characterised by the fact that you can work with more people of different profiles - in
this case with 50 participants in each of the five countries. We included three different profiles: potential,
existing and lapsed customers. The result was five MROCs to hear the opinions of ‘the many people’ in one
single project, on one platform.
The ‘research on research’ we did in the past taught us that for most participant profiles and research objectives
it is best to run a community in the mother tongue of the participants. Firstly, by doing so they will
discuss more and they will post in a more nuanced and emotional way. Secondly, our experience
showed that it is a must for the community to be moderated by a native who knows the language, the
local culture and the local market. All of this will lead to more relevant and to-the-point customer dialogues.
That is why we made the choice to set up a separate MROC in each of the five countries. This approach also
allows us to run the whole project at once and in parallel. While being in contact with participants via a local
moderator, you still grasp the advantages of having a central/global project team and content overview when
‘connecting the dots’ on a global level.
By working with a single master topic guide (which is adapted to the local reality and culture), interviewer bias is
partly eliminated. The project is run by a team of different people who work together both on country and global level.
11. 3 Transformational: doing things which were not possible before
Evaluating the Catalogue in an MROC
gives us the opportunity to work
with the same participants over a
longer period of time (in this case:
three weeks of 24/7 contact) in the
comfort and context of their own living
environment (home). In this project, we
followed the evolution in
perception and usage of the
Catalogue over time (before the
Catalogue arrived, the first flip-through
moments and the different reading
sessions). This is something that is
hard to realise cost-efficiently with
traditional methods.
Another advantage of having
more time with participants is
‘consecutive learning’. One can
build further on what one
learned previously. Or stimuli
material can be tested, adapted
and tested again in only a matter
of days with the same group of
consumers.
To obtain a holistic view on a
participant’s use and perception of the
Catalogue, a whole range of
research techniques was plugged
in into the community platform:
observational tasks, an online diary, a
collage tool, mini-surveys, creative
exercises, group discussions and even
implicit/emotion measurement tools.
Communities allowed us to blend
different research methods, giving
us depth and breadth in terms of the
insights we found.
12. The research team did not have a hard time convincing the
internal clients to make the switch, as it was clear that the time
had arrived for something different and better than
offline focus groups. The advantages of using a community
in the evaluation of the Catalogue (and the IKEA Catalogue
App) were very clear as well. The questions arising among the
business owners were more about the practical side of things,
e.g. how to choose and optimally combine the different
research options available on the platform, how to safely
distribute large numbers of Catalogues to different
parts of the world before the actual launch and how to
make sense of a big pile of information.
14. In order to bring fresh and unique insights to the table and to get a valid, clear and holistic view on people’s
perception and use of the 2013 Catalogue, we used the principle of ‘triangulation’ (Guin et al, 2012) to create a
true Fusion Research tool: investigating the same subject or issue from different angles and by doing so
create a more adequate and deeper understanding. We applied ‘triangulation’ on five different levels:
1 Data triangulation: by including both existing, lapsed and non-customers in the evaluation, we get data on
all three groups and a full view on how different receivers of the Catalogue react to it: what do they feel, think
and do?
2 Method triangulation: by combining different observational, qualitative (both interviewing and discussions)
and quantitative research techniques (both explicit and implicit measurement) we tackle the same issue from
different angles, leading to a holistic view of it.
3 Environmental triangulation: the mobile application used on the community platform allows
participants to not only take part on the main research platform from behind their PC, but to also
provide us with more personal and contextual information - in this case on how they use the
Catalogue on a day-to-day basis in their homes and over time.
15. 4 Theory triangulation: as we know that ‘people think less than we think they think’, it is important to take into
account both ‘System One’ and ‘System Two’ thinking when asking people to evaluate the Catalogue. That is
why we included both an implicit and an explicit test of the 2014 cover during the one-week reactivation of the
community.
5 Investigator triangulation: by asking the local community moderators, the global project team at agency
side, the client-side researchers, the internal client and the research participants to analyse certain parts of the
data, we made sure we got everything out of it and our final conclusions included different points of view from
people with different backgrounds. The local moderators keep an eye on aspects specific to the local culture
and market of their country, the global research team looks for the global consensus, the client-side
researchers and internal client frame the results within the business context and participants help us to close
the final blind spots we have. In this project we asked ambassadors from the five different country communities
to join forces in an English-speaking Global Room where a discussion took place concerning the cover test
during the reactivation, to see if and how it was possible to come to one cover for all countries. The view of the
participants helped us to see which cultural differences could be won over.
16. Applying the principle of ‘triangulation’ asks for more time and man-hours.
One needs to make the trade-off between cost and getting additional
understanding. In this project, it was a must to get a 360° view on how the
Catalogue was perceived and used over time. It was only by making use
of a ‘Research Community’ and applying the principles of
‘triangulation’to it, that we gained a fresh and complete view on
the evaluation of the 2013 Catalogue.
Next we give examples of each of the different tools
used to bring Fusion Research into practice
17. Phase 1: Evaluation of the 2013 Catalogue during three-week communities in five
different countries
During the first week the goal was to ‘meet the reader’
and understand the actual and aspirational
behaviour of the participants. Who is the reader
and what are his/her expectations towards the brand
and the Catalogue? To get there we conducted
amongst other activities a mix of an ethnographic task
to get a view of their house and insights into how they
live, a mood board exercise in which we asked the
members to map their feelings about the IKEA brand
and creative tasks on the forum of the platform (e.g.
Tell us ‘Your IKEA Catalogue story’).
After the Catalogue was dropped of at the
participants’ homes (in preview, two weeks before the
real global launch), we assessed the perception,
satisfaction and level of engagement the 2013
Catalogue evoked. First impressions and second
thoughts on particular aspects of the Catalogue were
researched (structure, pictures, stories, the Ikea
Catalogue App, etc.). It was important to understand
how the Catalogue offers both inspiration and
information to the reader and if that was done
to the right extent. Moreover, we wanted to get a
grip on the life cycle of the Catalogue. We did so by
adding three specific exercises:
18. A photo safari, in which we asked the participants to spot the Catalogue in
their house. We wanted them to imagine that the Catalogue could speak and
tell the story of his past, current and future life.
During the last two weeks of the community, a calendar tool was added to the
platform to get day-to-day insights in the usage and the emotions it evoked.
This was done in a private part of the community (participants could not read
the responses of others), which aimed at avoiding bias.
Finally the Ikea Catalogue App was tested by understanding the
expectations of these types of apps, asking them to use the app and to
evaluate it afterwards on its relevance, user-friendliness and discuss possible
improvements for the next version.
During the final week, we investigated whether the Catalogue was meeting expectations, giving us first clues
on the impact the latest Catalogue had on the brand perception and shopper behaviour. In other words,
what impact did it have on the business: was it attracting people to the shop, raising interest in home furnishing and
was there a positive impact on the perception of the brand, especially among lapsed customers?
19. Phase 2: Reactivation of the different communities (after eight weeks of inactivity) to
test the overall theme and the first cover ideas for the 2014 Catalogue
This was done in a two-step approach during one week:
A mini-survey (N=226, response rate > 70%) to test six options for the new cover (illustrating the theme)
Each participant randomly evaluated three different covers (resulting in at least 90 participants for
each cover tested). The questionnaire was a mix of explicit questions (fit of the covers with brand and
mission statements) and emotional/implicit measurement:
• Two-second test: each participant saw a cover for only two seconds. Afterwards, via an open-ended
question, the members of the communities were to write down what they saw, which emotions they
had felt and what they remembered. This was done three times for each participant in randomised
order, to exclude order effect. We were measuring ‘stomach impact’ here explicitly, though
spontaneously.
• Implicit measurement: each participant saw the same three cover pages for 10 seconds.
Afterwards, the members had to execute an ‘implicit measurement’ assignment where 15 emotions
were shown rapidly (one second) in randomised order. Participants had to press the space bar each
time they associated the emotion with the cover they just saw. Here we measured unconscious
emotional reactions to seeing the cover page.
20. By combining the breadth of the ‘stomach
impact’ on the X axis (the share of people who
associate a given cover with the emotion, via implicit
measurement) with the depth on the Y axis (for
those who have associated the emotion with the
cover, how strong is that association, i.e. how rapidly
was the association made). To make
interpretation easier, the scale of the Y axis
was reversed (= the higher, the faster) (see Figure
3). Results of these measurements were thrown
back into the discussion during phase 2, in order to
fully understand them.
21. Continuing with a discussion on the forum: in the survey, the participants were confronted
with different cover ideas
After two days they were asked which of the covers they still remembered. We used a tool in which you first
have to answer yourself, before the others can see your answer (which ensures that answers are not biased).
We did this to assess the ‘stomach impact’ on the mid-long term. After the spontaneous recall test, we
conducted the spontaneous associations test: upon showing the different covers, participants were
asked to share their initial thoughts, without prompting. Via this exercise we got an understanding of the
key associations made with each cover. Finally, by asking indirect questions (e.g. tell the story of the cover),
participants were motivated to elaborate on the indirect impact of the cover. Finally, the most remarkable
results of the survey were shared and the participants were invited to comment on them in order
to maximise our understanding.
‘Battle of the covers’ across the five countries in order to find creative ideas that would have
appeal across the globe
All participants of the five communities got the opportunity to take part in this discussion in English
in a central room. This way, cultural differences and local preferences were unveiled.
23. The previous paragraphs informed us about the fact that a
community provides one with automational,
informational and transformational benefits and that
it is a great tool to apply ‘triangulation’to. The
examples provided in the previous paragraph on how we
evaluated the 2013 Catalogue and pre-tested the themes
and cover for the 2014 edition show that it can be a handy
tool while creating a new Catalogue as well. A community
does not need to be ‘always-on’. You can perfectly align it
with business planning. In this case the communities were
used to draw learnings from the previous edition and
to get feedback on first ideas for the new one. We can
imagine situations later on in the creation process where we
might want to reactivate the communities for a short period
of time. Is that possible? Are participants indeed willing to
participate again and what does it take to encourage them to
do so?
The answer to the first question is ‘yes’: almost all
participants from the communities in the US, Poland
and Germany took part again. For Italy (78%) and
China (70%), the reactivation was a bit less
successful, although this was neither expected nor
communicated at the start of the community.
Cultural differences in commitment are
probably at the basis of the lower numbers in
both countries. For all the questions on the different
communities, we had at least 30 posts, which is what
we needed to reach our saturation effect (Schillewaert
et al, 2011). Across the five communities we had
2,807 interactions in a week’s time, which is 38%
of the 7,261 we generated in the initial three weeks.
This is perfectly in line with what one may expect.
24. By making the participants feel part of the company as real ‘consultants’. Before the first
phase of the project a kick-off was organised in a 30-minute chat session in smaller groups, during
which we explained who the client was, what the goal of the project was and what was in it for them.
During the first phase, participants received weekly newsletters on the progress of the project
and what the company was learning from the discussion.
How did we get to these results?
After the first three weeks and the workshop at the company
side the participants got an update on how the debrief in
Sweden went and pictures from the office were the Catalogue is
created and a word from the team behind it. It is by doing so
and by keeping your promises in terms of incentives that
you create a strong relationship with the members.
26. The fundamentals of the community approach work on a global scale. Just like the brands
we are working for, we need to localise our way of working from country to country. In
order to fully understand to what extent localisation of our methodology is required, we conducted
several studies with moderators from our ‘Global Community Moderator Network’
(recruited in 30 different countries) and with local research participants. This way we co-created
best practices for the different markets we are operating in. We found that it is important to
adapt your community to the local culture on five aspects. The direction the adaptations are
made in can almost always be explained by the work of the Dutch academic researcher Geert
Hofstede and his five dimensions to explain cultural differences between countries. Next we
explain the different dimensions in more detail and we also add some striking examples.
27. 1 Reason to participate: intrinsic vs. extrinsic motivation
From the first e-mail invitation for joining the
community onwards, it needs to be clear
what is in it for the participants. In
almost all countries, the main reason to
participate is the possibility to have an
influence on the future of a brand or a
product.
We also noticed that some countries are more
extrinsically motivated than others. This is especially
the case in the US and in Eastern European countries, but
for different reasons. Americans consider it normal that
there is a payment to reward performance. In most
Eastern European countries on the other hand an
(monetary) incentive is perceived as a nice extra on top of
their monthly income. Furthermore, in Poland, it is a must
to gain the ‘trust’ of the members - trust in the fact that the
agency or company behind the community will not harm
them in any way and also in the fact that they will really
get their incentive. The preferred type of incentive
differs from country to country. It is an illusion to
think that ‘PayPal fits all’. In Asian countries such as
China, the intrinsic part is important: they like to be
connected with aspirational brands and share
their wisdom.
28. 2 Conversation guide: empowering vs. directive
A different culture also means different
attitudes and values, leading to a different
way of reacting to certain questions, tasks and
exercises that moderators want participants to
perform. Some cultures, for instance, love to share
a lot of details about themselves and their lives.
Others prefer talking about the group, which is
considered to be a safer option. See it as a
projective technique to let people talk about
their own situation, free of any pressure. The
same holds for co-creation exercises. It is not a
given in every culture that people are used to
taking initiative. They feel better when they are
only asked to give feedback about what already
exists. It is important to map the country that
one is working in on those two axes (‘me’ vs.
‘we’ and ‘feedback’ vs. ‘co-creation’) and to adapt
the way of writing and (re)mixing topics for
the conversation guide to it.
29. 3 Role of the moderator: facilitator vs. authority
One does not only need to adapt the way of
inviting and incentivising the members, the
medium of data collection and the nature of the
topics in the conversation guide. The role of
the moderator is also perceived differently
from one country to the next. In Italy a
moderator needs to facilitate and start the
discussion. His/her role lies more in the
background.
But it is also expected from the moderator that he/she
is steering the discussion in the right direction when it
is going off topic. In Poland on the other hand, the
moderator needs to be strict and almost literally direct
the members to the next question or task they need
to look into. At the same time the moderator in Brazil
is a social peer, he/she needs to be a formal
professional in China and a like-minded person to
exchange wisdom with in India. It is crucial to
know and manage all these different
expectations when running (multi-country or
global) ‘Consumer Consulting Boards’.
30. 4 Gamification: playful vs. serious
Adding elements of ‘gamification’to the
community brings more richness to the
table. In our research-on-research among our
moderators, we learned that the level of and
the intensity with which you gamify your
‘Consumer Consulting Board’ need to
differ between countries. In Germany, for
example, it is wise to limit it to a minimal level as
it is culturally less accepted.
31. The elements described on the previous slides show that in multi-country projects you need to
start from a master conversation guide, which is important to make sure that there is a uniform
way of working and that you exclude the effects of interviewer bias. But adaptations both in
content and style of the topics and in the way of moderating will be amongst other key
elements in making the community a real success. Furthermore, it is wise to plan several
debriefs between the different local moderators facilitated by the global research team: to
challenge each other’s conclusions, let them go back to the results of their own country and come
back with deeper and richer understanding in several iterative loops.
33. Only the business owners were following the
community closely, although it should be convenient
as no travelling is required and one can take a
look at the discussion on the community when
one feels like doing so. As a researcher there is a
clear need to give intermediate updates to the different
stakeholders, as they do not follow the discussion
spontaneously and there is a lot of information to
digest. Moreover, this gives the business owners the
opportunity to finalise the topic guide for the upcoming
days: going deeper into certain elements that are
really interesting or pushing the discussion in a
new direction.
Previous ‘research-on-research’ (De Ruyck, 2011) and
our experience during this project have taught us that
confronting stakeholders with real stories by real
people is very impactful as an illustration of the
main conclusions. They become alive. It is also an
advantage that stakeholders can go back to the
community based on the final report and read
exactly what and how customers put it during the
community. Finally, it is great that a community can be
reopened when you need it, perfectly in line with
the business planning.
The research results have led to significant changes to
the Ikea Catalogue App. We now know that the new
concept of the Catalogue was a big step into a new
and right direction and improvements will be
made for the 2014 edition, based on the research.
35. The power of ‘Research Communities’ as a methodology lies in the fact that you have the ability
to work with more people, over a longer period of time, and that you get to know a lot
from all kinds of different angles by combining different research tools and methods. The
latter especially is still untapped potential in most communities. This case study describes how
fusing observational, qualitative and quantitative research methods lead to a deeper
understanding and new insights. The case also demonstrates that ‘Structural Collaboration’ with
consumers over time is valuable and leads to more impactful communication tools.
37. De Ruyck, T. et al (2010). ‘How Fans Become Future Shapers of an Ice-cream Brand’,
Proceedings ESOMAR Qualitative
Guin L., Diehl D. and McDonald D. (2012), Triangulation: Establishing the validity of
qualitative studies, IFAS
Schillewaert, N. et al (2011). ‘The Darkside to Crowd-sourcing in Online Research
Communities’, CASRO Journal
De Ruyck, T. et al (2011). ‘Engage, Inspire, Act: 3 Stepstones towards Developing more
Impactful Products’, Proceedings ESOMAR Congress
38. Tom De Ruyck
Head of Consumer Consulting Boards
InSites Consulting
Pieter De Vuyst
Senior Research Manager
InSites Consulting
Frédéric Gennart
Global Market Research Consultant
Inter IKEA Systems
Frank Naessens
Senior Research Consultant
InSites Consulting