How cars are really bought full deck esomar research 67596 09 desor-ellis

CONGRESS 2012 Copyright © ESOMAR 2012
HOW CARS REALLY GET BOUGHT
BEYOND THE PURCHASE FUNNEL - NEW INSIGHTS FROM DIGITAL ETHNOGRAPHY
Neel Desor • Rob Ellis
ESOMAR Office:
Barbara Strozzilaan 384
Eurocenter 2
1083 NH Amsterdam
The Netherlands
Tel.: +31-20-664 21 41
Fax: +31-20-664 29 22
Email: customerservice@esomar.org
Website: www.esomar.org
Publication Date: September 2012
ESOMAR Publication Series Volume C12
ISBN 92-831-0260-6

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system of any nature, or
transmitted or made available in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without the prior written permission of ESOMAR. ESOMAR will pursue copyright infringements.
In spite of careful preparation and editing, this publication may contain errors and imperfections. Authors, editors and
ESOMAR do not accept any responsibility for the consequences that may arise as a result thereof. The views expressed by
the authors in this publication do not necessarily represent the views of ESOMAR.
By the mere offering of any material to ESOMAR in order to be published, the author thereby guarantees:
• that the author - in line with the ICC/ESOMAR International Code of Marketing and Social Research– has obtained
permission from clients and/ or third parties to present and publish the information contained in the material offered
to ESOMAR;
• that the material offered to ESOMAR does not infringe on any right of any third party; and
• that the author shall defend ESOMAR and hold ESOMAR harmless from any claim of any third party based upon the
publication by ESOMAR of the offered material.
Published by ESOMAR, Amsterdam,
The Netherlands
Edited by: Deborah S. Fellow
COPYRIGHT

ESOMAR is the essential organisation for encouraging, advancing and elevating market research worldwide.
With more than 4,900 members from over 130 countries, ESOMAR’s aim is to promote the value of market and opinion
research in illuminating real issues and bringing about effective decision-making.
To facilitate this ongoing dialogue, ESOMAR creates and manages a comprehensive programme of industry specific and
thematic events, publications and communications, as well as actively advocating self-regulation and the worldwide code of
practice.
ESOMAR was founded in 1948.
ABOUT ESOMAR MEMBERSHIP
ESOMAR is open to everyone, all over the world, who believes that high quality research improves the way
businesses make decisions. Our members are active in a wide range of industries and come from a variety of professional
backgrounds, including research, marketing, advertising and media.
Membership benefits include the right to be listed in the ESOMAR Directories of Research Organisations and to use the
ESOMAR Membership mark, plus access to a range of publications (either free of charge or with discount) and registration
to all standard events, including the Annual Congress, at preferential Members’ rates.
Members have the opportunity to attend and speak at conferences or take part in workshops. At all events the emphasis is
on exchanging ideas, learning about latest developments and best practice and networking with other professionals in
marketing, advertising and research. CONGRESS is our flagship event, attracting over 1,000 people, with a full programme
of original papers and keynote speakers, plus a highly successful trade exhibition. Full details on latest membership are
available online at www.esomar.org.
CONTACT US
ESOMAR
Eurocenter 2
Barbara Strozzilaan 384
1083 HN Amsterdam
The Netherlands
Tel.: +31 20 589 7800
Email: customerservice@esomar.org
ABOUT ESOMAR

INTRODUCTION
Buying a car is the second largest purchase most of us make, after a home. There are many well-established brands, and
huge marketing budgets devoted to persuading us to choose one marque or model rather than another. There is also a
range of specialist magazines advising us about choice and informing us of news, and television shows about cars get high
ratings.
There has also been a substantial investment in market research, from initial concept research, car clinics and ride and
drive sessions, customer satisfaction and post purchase surveys, brand evaluations and purchase intent and ad tracking
studies.
Despite this investment in research many in the industry still felt there was more to be learnt about how consumers really
chose cars, and how they found their way on or off the shortlist, and at COG and Haymarket we often encountered
scepticism about whether research really addressed key brand choice issues.
Haymarket is an international media company with a long and distinguished heritage in the automotive sector. Amongst
other activities, it publishes the world’s oldest car magazine, the most influential UK buyer’s guide, and curates the largest
UK online enthusiast community.
Although Haymarket has extensive real world experience of this market, and an ongoing dialogue with buyers, there is
always more to learn. Consumer behaviour is constantly evolving, the media landscape is changing daily, and the pace of
product development from manufacturers is unrelenting.
Automotive is a well-established sector, and with this familiarity comes perceived wisdom and conventions. In their day to
day shorthand, people in the industry still talk about funnels, shortlists and consumer motivations in the same way they did
20 years ago, laced with assumptions that buyers are entirely logical, rational and linear. We knew from our journalists,
readers and users that many of these conventions did not reflect actual behaviour.
Haymarket has a reputation in the sector for commissioning and publishing original research on cars and car buying. The
objective was to enhance its hard earned reputation as thought leaders in the automotive media world, principally amongst
clients – car companies and their media agencies, but also to extend this to other communication agencies and the broader
automotive industry. This objective would deliver important commercial benefits for business, which are outlined later.
In late 2010 Haymarket approached COG Research to design a new research project, which would add to the body of
knowledge, challenge existing beliefs and create new evidence, and stimulate discussion.
This paper describes the project that grew out of that discussion, which was a new way of looking at the car buying
process designed around capturing the process, the touchpoints and the trigger events that eventually led to purchase,
rather than being based on consumer self report.
The main objective of this research was to provide genuinely new and useful insight into how cars were being bought, and
how the purchase process really worked. The secondary objective was to use such new insight to provide opportunity for
a dialogue between Haymarket and car manufacturers and agencies.
Haymarket and COG both had extensive experience of working with car manufacturers and their agencies, and this
allowed us to identify the areas where existing information or theories were widely felt to be inadequate or unhelpful. The
framework of Behavioural Economics was attracting wide attention as a new insight into how purchase decisions were
HOW CARS REALLY GET BOUGHT
BEYOND THE PURCHASE FUNNEL - NEW INSIGHTS FROM DIGITAL ETHNOGRAPHY
Neel Desor • Rob Ellis

being made, and new learnings from neuroscience about explicit and implicit processing also helped inform the discussion
about purchase decisions.
The brief for this research was to aim to identify new and more reliable evidence on how the car purchase decision
purchase was actually being made by consumers, as opposed to how they self reported their decision process (typically in
post purchase surveys). Work started from the hypothesis that consumers (or at least car buyers) were often unreliable
witnesses to the multiple touchpoints, decisions, feelings and other triggers that led toward a final purchase decision. A
number of writers around purchase decisions such as Ariely (2008), Harford (2005) and Levett (2006) had highlighted
instances when consumers revealed behaviours didn't match their claimed rational criteria, especially in the area of
purchasing. This paper is not intended to explore the reasons for that unreliability in detail, but it became clear that there
were two barriers to accurate reporting of complex decision patterns:
1. The complexity, detail and length of the new car search process meant that it was often difficult to recall all the
touchpoints and elements within the process accurately (we found an average of 14 cars on the shortlist and in some
cases over 50).
2. Our own views on how we make decisions and what is important to us naturally create a context for the recall of a
decision process, so that while we may not consciously choose to edit our recall, we will be more likely to recall
incidents or evidence that fits how we think we made a choice. These tendencies have been discussed at length in
studies about confirmation bias and the availability heuristic (Sutherland et al, 1992).
THE RESEARCH DESIGN AND METHODOLOGY
The discipline of ethnography identifies that the primary task is to identify what people actually do, and then to go on to
capture the reasons they give for doing it, before conducting analysis or attempting to interpret their actions (LeCompte &
Schensul, 2010).
FIGURE 1, ETHNOGRAPHY IN NORWAY AS IMAGINED IN THE FILM ‘KITCHEN STORIES’ (2003)
This was exactly what we wanted to do in the car market, but there were formidable difficulties. The decision process was
known to take a long time (average estimate over three months from decision to choose a new car to placing an order,
according to industry research), and the process was often passive rather than active (seeing ads, meeting friends with
new cars, chance trips in cars, seeing cars on road, etc.). Conventional ethnography where the researcher literally
observes the action would be completely impractical.

Our solution to this was to develop a light touch form of ethnographic study called digital ethnography, which allowed
respondents to record any relevant experiences or touchpoints on a mobile device. We had previously used this approach
in a study of young people via smart phones and had found that the uptake and speed of response was much higher than
for conventional web-based research.
Because most people had their mobile phones with them all the time, and regularly used the phones for communication
with friends and family, we believed it could be a practical and engaging medium for running a simple diary type study, and
for capturing thoughts in the moment. It proved extremely effective, and our experience was that responses to questions
or prompts were both quickly given and sufficiently detailed. Where we needed more detail we simply texted back
questions (using a web based text platform and dedicated researcher) and when we did this we found levels of
involvement increased further.
This approach formed the backbone of the new car buyer study as outlined below. We used face to face recruitment to
assure ourselves of respondent quality and to assure them of our seriousness (they needed to trust we would reward
them over a prolonged period). We then carried out a phone briefing and an initial online survey, and then allowed them to
contact us whenever any interactions took place. The length of the interaction period was dictated by how long they took
to choose a car, and the criteria for success were that they placed an order (car delivery might involve further waiting).
FIGURE 2, THE RESEARCH PROCESS
As well as waiting for them to let us know about any related events (which they did regularly, with an average of 110
messages per respondent and some respondents messaging us 12 or more times a day at busy periods) we also
maintained contact from our side. We invited them to complete a weekly online diary that captured in more detail any
activities during the week, and the latest version of the shortlist. We messaged them regularly to thank them for their
comments and in some cases ask for follow up detail, and to remind / check on respondents who had been quiet for a
week. We invited them to themed surveys on personality traits, specific activities such as test-driving, brochures and online
searching (these were only served later in the process once they had either taken part in the activities themselves or had
decided on purchase). We carried out video interviews with all respondents at least once during the process and again at
the end of process after they had chosen a car or decided to remove themselves from market. We also carried out a
follow up online study both among all participants and a matched fresh sample of new car buyers to use as a control
(identifying if our close following of their behaviour had obvious effects on them vs. a control audience). As we knew
exactly which car all buyers eventually chose, we were able to identify the value of their combined purchases as slightly
over £3.5 million.
This approach meant that the research was essentially built around each individual’s journey, with the research stages
capturing and mirroring the individuals’ progress through the car buying process.

RESULTS AND DISCUSSION
It became clear that there were a number of buyer segments in terms of the typical length of process and the depth to
which they engaged with it. This segmentation has since been validated by more quantitative research, and has proved
extremely helpful to manufacturers planning communications and promotional strategies.
The three segments are shown in figure 3, which shows the proportions we found in the study. They were derived from
actual behavioural data (how long the search was, how many cars shortlisted, how much active research undertaken),
psychological traits research (the importance of information, risk tolerance, importance of peer opinion, derived from tests
developed by DeMarchi and Hamilton) and self reported data (interest in cars, knowledge about and confidence about
cars). The segments were also analysed by the touchpoints and influences that were key for each group, indicating the
role played by word of mouth for speedy choosers, for magazine reviews among maximisers, and the extent to which
enthusiasts also influenced others.
FIGURE 3, THREE TYPES OF BUYER BASED ON BEHAVIOUR
Other areas in which the longitudinal approach gave us real insight were the influence of advertising and test drives, which
respondents in pre-choice research tended to under and over-estimate respectively.
Test driving both proved to be less frequently undertaken than expected, and yet to be a deal breaker in terms of cars
falling off shortlists. If as part of the search process there was a test drive requested that could not be provided, this
resulted in the car being taken off list in nine out of ten cases. Clearly, failure to provide a test drive when asked was a
major negative for the consumer. However, absolute numbers of test drives were lower than predicted, and two-thirds of
sample took only one or no test drives and only one in three made any kind of comparison.
FIGURE 4, ONLY ONE IN THREE COMPARES AT TEST DRIVE
We also found that for those taking one test drive this occurred late in the process and in over 80% of cases was followed
by a purchase. It seemed this wasn't so much a test drive as a final confirmation. This has suggested both that every effort
must be made to ensure a car is available if requested (as the up side is a likely sale and down side is a lost sale), and that

broader test driving programmes could help drive conquest sales. The important evidence here was the exact place that
test drive requests came in the individual’s purchase journey, and the subsequent purchase pattern they exhibited. Over
time the specifics of how a request is handled may get lost, as we found when we followed up on respondents who had
de-listed cars following failure to supply test models. At the time the lack of suitable car was clearly identified as the
trigger for de-listing, but by the wrap up survey it had been re-expressed as a general dissatisfaction with the dealer, and
the test drive incident was only recalled on prompting.
One issue that we faced in reporting this complex longitudinal data was how to chart the complexity of each respondent’s
interactions over time, as it became clear that following each individual’s story was crucial to understanding their journey
as well as the overall data. This was a key issue as we knew we would eventually need to show manufacturers the stories
related to their buyers (or lost buyers) in a compelling and easy to understand way. We looked at a range of ways of
displaying information, studied Tufte (1983) and eventually adapted the river visualisation as shown in figure 5, as it could
display key information on an individual’s journey. This included the start and end points, number of cars on the shortlist at
each week of process, actual list of cars shortlisted, and summary triggers for adding and subtracting cars or other key
events. The latter were hyperlinked to more detailed data in the client version.
FIGURE 5, THE SHORTLIST IS A LIVING THING
The study yielded over 70 of these stories, some over shorter and some longer time frames, and some even more
complex than this one. Close study revealed intriguing detail for our automotive clients that we found repeated in many
other purchase journeys. Some salient points are:
• The number of cars on the shortlist first grows, and only then starts to decline (which makes perfect sense, but
contrasts with the funnel notion, which tends to lead us to think of the long list leading to the short list. This expansion
then contraction of the shortlist was frequently seen in the study.
• The actual cars on shortlist are not by any means within a coherent segment. An Audi A2 may be considered, along
with a Hummer and a BMW 5 series, at different weeks of the journey. Again, we saw this cross segment
phenomenon frequently occur. However, in post purchase surveys we found the proportion of ‘cross segment’ cars
mentioned to be substantially lower, suggesting that we may be unconsciously editing our recalled short lists.

A much simpler purchase journey is shown in figure 6.
FIGURE 6, FIRST TIME BUYER
This was a classic example of a speedy chooser, and it is interesting to note that the car eventually chosen was not on the
original shortlist. They made their original list by consulting a magazine and thinking of friends’ cars. Only after their initial
list did they remember the friend with a Mini, and recall their interest in the car. Once they had visited the dealership they
became clear this was the car for them.
We have found these river charts to be extremely compelling and thought provoking in discussion with manufacturers and
agencies – they highlight the process of choice on a case history basis that allows people in the business to think again
about how cars are actually getting bought, and move beyond the attractively simple model of the purchase funnel.
FIGURE 7, THE CLASSIC PURCHASE FUNNEL MODEL

HOW HAS IT WORKED? THE CLIENT’S PERSPECTIVE
The brief for this research was to provide new and credible evidence of how car buying really took place, and for this
research to gain interest for itself and the client in the motor and media industries. To achieve this, the research had to be
both clearly and intuitively novel and credible, so that Haymarket (and COG) could secure a hearing among two industries
that are used to dismissing PR exercises based on limited research. Based on the criteria the project has delivered
benefits to Haymarket on a number of levels.
The findings have been presented to 26 marketers at 12 leading car companies and over 100 planners and buyers at
major media and advertising agencies. These people are collectively responsible for the vast majority of advertising spend
in the sector, and constitute over 90% of our display advertising revenue.
The project has undoubtedly given us a higher profile amongst our clients – the lifeblood of our business.
It has opened doors, giving our advertising team high levels of access to new and important people. It has also made our
advertising team proud of where they work and what they do, and has taken their conversations beyond day to day
trading.
The research was also presented to a large audience at the Society of Motor Manufacturers and Traders (SMMT) annual
conference. The SMMT is the pre-eminent trade body for the car sector in the United Kingdom. The SMMT presentation
was accompanied by live tweets to Haymotoring, our industry twitter followers.
A key objective of the study from Haymarket’s point of view was to enhance our reputation for thought leadership
through commissioning innovative and insightful research: the feedback we have had from the media and
automotive industries suggest this objective has been fulfilled.
The number of follow up questions and requests for data is testament to the fact that people are listening and that they are
engaged. The president of the British IPA (Advertising Industry Body), Rory Sutherland, said “I have always instinctively
felt there was quite a bit of friction in the car-buying process - partly because buyers were expected to behave in ways
that simply do not accord with human nature or what we know about decision making - and that the conventional funnel
model was not just a simplification but actively misleading. This research gives me that rare but gratifying feeling of having
my instincts confirmed.”
The research is, in itself, a brilliant piece of client marketing.
HOW HAS IT WORKED? THE RESEARCHERS’ PERSPECTIVE
Going into the project, there were a number of concerns about practical and theoretical matters, some of which were
resolved in the course of the research. Subsequently we have conducted a number of exercises in similar fields (including
media viewing, consumer good and food purchasing, and alcohol consumption) that have benefitted from the lessons
learnt in this pioneering study.
Key concerns at the time of commencing the study were:
1. Will a robust sample continue with engagement with the process over an extended period, or will there be problematic
levels of drop out or reduced participation?
2. Will the participation be representative of the actual market or will it skew to different demographic, car segment
buyers or user types?
3. Will extended participation in the exercise change the dynamics of choosing behaviour so that true behaviours are
distorted?
4. If this study yields new information about how people choose, how will we know whether to trust this or past
research evidence?
The first concern, about participation rate, was one of our greatest worries in setting up the study. We knew that
conventional consumer panels had typical attrition levels, but this was hardly a conventional panel. We proposed to
incentivise participation on a monthly basis, and to reward active participation through a weekly prize draw across the
panel, and this seemed to be effective in terms of supporting high levels of participation. In hindsight, the fact that
respondents were being recruited to report on a task they were in control of (their car purchase) added to the sense of
willing participation. Eighty-eight percent of the original respondents recruited to the ethnographic study remained
involved, although not all bought a new car, as shown in figure 8.

FIGURE 8, BUYING STATUS AFTER SIX MONTHS: FOUR OUT OF FIVE POTENTIAL BUYERS SUCCEED OR
STILL SEARCHING
Most actually did buy a car or were still in the process of looking, some decided to buy second hand instead and others
dropped out, mainly because they had decided against a car purchase at this time.
This level of retention has been repeated in subsequent studies, and when we have spoken to respondents about their
experiences (at the end of the research process) they offer two reasons for their willingness to continue taking part: it was
easy to do (the simple text / app based interface) and it was interesting and rewarding to feel that their experiences were
being heard (the feedback system from project managers). The incentives, while obviously welcomed, were seldom
mentioned explicitly as a key reason for continued participation.
This high level of retention also addressed the second concern: as we had set out to recruit a representative sample in
terms of demographics and potential car segment, the sample remained representative with no differences in dropout level
among the sample segments. In terms of chooser types, it became clear that length of time to choice was a key
segmentation element, so the slow choosers were disproportionately represented in the ‘still looking’ category. As a result
we targeted these individuals and persuaded some to stay in touch with us, so that we could get a more complete picture
of their eventual car choice. Most eventually did choose a car, although we know of some who were still looking nearly a
year later.
Did extended participation change the dynamics of car choosing? We analysed the nature of the panel’s feedback in the
first and last month of their search, and while there were more dealer interactions toward the end (as you would expect),
the proportion of other touchpoints being mentioned remained similar, with advertising, noticing cars and conversations
with friends and family remaining at similar levels.
There was a more obvious effect which we have termed the internal cycle of choice, where level of feedback overall,
adding and dropping cars from shortlists, visiting or contacting dealers and websites and other symptoms of active
choosing all tended to move in tandem. This would suggest that external factors in our lives dictate the amount of attention
we give to the car choice, rather than the research process artificially heightening it.
Our contact with respondents also showed that work, holidays, family and other life events often explained a period of
inactivity in terms of the car search, and that at some point they became more active again. In the wrap up study we
asked people if they felt that participation had changed their behaviour and over 90% said they didn't believe it had. For
those who did report a change it was more likely to be in terms of heightened awareness of advertising or their own
observations, rather than a perception of their behaviour having changed.
“Once I started doing this, I realised I was paying more attention to the car ads. But I think that is because I am looking for
a car that I am noticing the ads more” M 35 – 44
“I have noticed that I am paying more attention to cars, if I see a new one I like the look of” F 45 - 54
So do we trust this new evidence or the classic surveys on how car choice is made? At the beginning and end of the
process, we carried out an online survey with all participants and at the end of the process also recruited a matched larger

sample of recent new car buyers. The intention of this was to compare the observations we made and the reports from
respondents in the moment with their reflections in hindsight on what would or had influenced them.
Some reports were consistent and some disparities were striking: we have shown the key differences in table 1.
TABLE 1
Tend to over-estimate likelihood or
importance of this factor
Tend to be accurate about this
factor
Tend to under-estimate likelihood
or importance of this factor
Test drives – we assume we will be
more likely to take than we do
We are usually accurate in recalling
how many we take
Marque choice - we tend to be quite
accurate about predicting what will be
on shortlist
Advertising – we under-estimate how
much influence it will have on short-
listing, and we significantly under-
estimate the effect it did have on cars
being short listed
Best deals – we tend to assume we
will be more deal driven than we
actually are
Budget and finance – we tend to be
quite good at estimating total cost of
deal
Length of process – we tend to
assume we will be faster choosing
than we actually are
Some of these mismatches can be explained by attention and memory, or by behavioural economics. For example, we
found that we were more likely to be influenced by trade in price on the existing car than by a discount on the new car,
which fits with the concept of loss aversion (we hate losing money more than we like gaining it). Kahneman, Knetsch, &
Thaler, (1990)
The under-estimate of advertising impact on short-listing can be explained by two factors; the tendency to forget the first
trigger to a complex set of behaviours, where an ad may (and did) trigger consideration of a car, then the online research
about it, then visit to dealership and / or discussion with owner friends, and then perhaps a brochure request and reading
a review, with in some cases a final test drive. When we interviewed respondents about why they had not cited advertising
as one of the triggers to consideration (when we knew from their daily feedback that it had been a trigger) the usual
response was ‘oh, yes, I did see an ad now you remind me’. In some cases there may be a conscious belief system that ‘I
am not influenced by advertising’ but in other cases it was more plausible that the original ad simply had been forgotten as
a trigger. This may also be explained by availability, in that we over-estimate the influence of things we have put effort into
(and are thus more available to memory) vs. things we simply happen to see.
One of the benefits of this methodology was to follow up differences between expectations, actual behaviour and recalled
behaviour. In almost all cases, when the respondent was reminded of actual behaviour they agreed that that was in fact
true, but that they had expected or remembered the process as being different, and the expectations and the memory
were what they recalled in the survey interviews.
Given this, we believe the findings from the behavioural stage of the study are accurate, and moreover help us to
understand where and why expected behaviour or recalled behaviour might be inaccurate as a record of the actual
process. It ties in closely with the theories underpinning behavioural economics, which posit that we ‘often make poor
choices – and look back on them with bafflement’ (Thaler and Sunstein, Nudge, 2008).
IN CONCLUSION
This study set out to provide better insight into how people made an important purchase decision - their next new car. The
belief that we would learn more about the real decision from following people through the process was completely
vindicated, and we learnt much more than we could have expected from a simple post purchase survey.
The accurate information about touchpoints and the purchase process meant we could build more accurate and useful
segmentations, and we were able to track the real effect of ads (greater), influence of test drives (less, but with huge
potential), importance of word of mouth, dealers and online research.
The purchase funnel model was replaced with a new image of the organic and evolving process – an image that car
marketers have found useful and illuminating.
We believe this project has wider implications for anyone conducting purchase process research or planning to target
buyers of higher value goods: it has also allowed for evidence of how behavioural economics theory is at work in an actual
decision process.

The reaction to this study (which was intended by its commissioning client to be widely shared) has been very positive in
the marketing community: the approach taken and the findings generated have been of great interest to a wide marketing
and research audience, and there have been many requests for the presentation to be shared.
Other studies we have undertaken over the last year show how ubiquitous smart phones have become, as information and
communication resources, with shopping, entertainment and information apps enjoying wider uptake. We are already
conducting new studies using this methodology and find respondents very comfortable with digital self-reporting: we
expect this strand of research to become much more widespread in future and believe that it will lead to better
understanding of true consumer motivations and behaviours.
REFERENCES
Ariely, D, 2008, Predictably Irrational, Harper Collins
Harford, T, 2005, The Undercover Economist, Oxford
Margaret Kahneman, D., Knetsch, J., & Thaler, R. 1990. Experimental Test of the endowment effect and the Coase
Theorem. Journal of Political Economy
DeMarchi, S and Hamilton, J, You are what you choose, 2009, Portfolio
Diane LeCompte, Jean J. Schensul, 2010, Designing and Conducting Ethnographic Research, Rowman and Littlefield
Levett, S, Dubner, S, Freakonomics, 2006, William Morrow
Sutherland, S, Irrationality, 1992, Constable
Thaler, R and Sunstein, C, Nudge: Improving decisions about health, wealth and happiness, 2008, Yale
Tufte, E, The Visual Display of Quantitative Information, 1983, Graphics Press
THE AUTHORS
Neel Desor is Commercial Planning Director, Haymarket Consumer Media, United Kingdom.
Rob Ellis is Director, COG Research, United Kingdom.

How cars are really bought full deck esomar research 67596 09 desor-ellis

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to How cars are really bought full deck esomar research 67596 09 desor-ellis

Similar to How cars are really bought full deck esomar research 67596 09 desor-ellis (20)

More from dkcvoom

More from dkcvoom (6)

Recently uploaded

Recently uploaded (20)

How cars are really bought full deck esomar research 67596 09 desor-ellis