Highlights from ESOMAR Congress: Day 1, afternoon. Covered by Erika Harriford-McLaren and bloggers Tom Ewing, Anna Peters, and Kathy Joe.Celebrating Business: Using Insights to Drive Real Business Growth. - Tom Ewing
This session wasn’t billed as a celebration of new methods, but that’s what it turned into. It turns out – as if we doubted it! – that insights from co-creation, MROCs and mobile touchpoint data can inform and inspire as well as any presented in ESOMAR’s 65-year history.
Heineken and InSites teamed up for the first of three excellent case studies, focusing on Heineken’s desire to understand clubbers – and to bring that understanding to life by designing a “concept club” to perfectly meet clubbers’ needs. This club became reality at the Salon di Mobile in Milan this April, and Tom DeRuyck of InSites and Henk Eising of Heineken walked us through a parade of innovations. There was a touchscreen beer-ordering system to remove bar queues, concierges to ferry clubbers to after-parties when the night ended, and numbered shelves partygoers could rest beer on while they went to dance – this last would have saved me a lot of money back in my more nimble days.
All these innovations, of course, were based in consumer insights – 120 young clubbers with an interest in design (and beer!) that InSites had recruited for a three-week qualitative community exploring the emotions, culture and experience of clubbing. Their insights were fed – via an elegant website, nightlifejourney.com – to the up-and-coming designers Heineken and InSites recruited to build the club. The joy was in how beautifully InSites and Heineken activated their findings: it’s rare and refreshing to find a research project with such a strong aesthetic sense, and it demonstrates yet again how research can inspire creativity rather than limit it.
From the achingly hip to the – well, not quite so fashionable: the 120-year old Ladies Home Journal brand, whose publisher Meredith turned to Communispace for help with a makeover. Communispace’s Manila Austin opened with a moving story about her family’s experiences with her grandmother’s Alzheimer’s in the 1970s. It served as a gentle rebuke, too, to keynote speaker Sherry Turkle’s rejection of digital connectedness – back then families had plenty of time for face-to-face conversation but no ability to share information, or find others going through the same thing.
Back then a magazine article might have served as an authority and lifeline, but it’s a different world now, and authority has diffused across social networks. Meredith decided to outsource the Ladies Home Journal content to its readers, drawing on blog content, comments and Facebook stories to refresh the magazine’s image and style, while using traditional journalistic skills of sub-editing and fact-checking to create a compelling package. Central to this was a Communispace-built community of 300 women which both inspired and evaluated the new outlook. Austin talked about how online communities offer an intimacy and honesty which traditional research lacks, and video of enthused participants backed her up.
The community and its more honest tone directly inspired the magazine, Britta Ware from Meredith told us, liberating it to be funny and more open about its reader’s lives. And in attitudinal terms the makeover was a great success, with a 500% increase in Facebook likes and more than 1500 readers applying to become “reader-editors”. I’d have liked some harder metrics too – how circulation and advertising responded to a more radical publication – but as a study in how a research community can be placed at the heart of a brand this was a winner.
Finally, COG Research and Haymarket Consumer Media talked about how the car buying process really works. This was one of the most entertaining and insightful presentations I saw all day, gleefully exploding the automotive sector’s myths about considered consumer decisions.
Car buying and its research is a hive of post-rationalisation – because a car is a big ticket purchase, it logically follows that the decision to buy one must be involved, long and serious. And it suits everyone concerned – from consumers to manufacturers – to believe this. But is it true? The COG research, entertainingly presented by Neel Desor of Haymarket and COG’s Robert Ellis, took an empirical approach, carefully recruiting 80 buyers and following them through their decision making journey by way of mobile touchpoints. Every time they had a conversation about a car, saw an advert, went for a test drive, researched online, or did anything else to do with the decision, they told the researchers.
The results were fascinating. Rather than a gradual narrowing down, purchase shortlists remained in wild flux throughout the process, sometimes doubling halfway through. New brands would appear throughout, and often the eventual purchase was very quickly decided. The study junked cherished notions of the importance of test drives as comparative decision makers – two-thirds of buyers only took one, or none at all. Finally, while brand trackers downplayed the importance of advertising, touchpoint analysis revealed it was the strongest predictor of the eventual decision.
This wasn’t presented as a ‘behavioural economics’ study, but it ended up confirming most of the predictions behavioural economists would have made – our car buying decisions are often as irrational and rapid as any other purchase. My favourite finding was that while a third of buyers are indeed “benefit maximisers” – carefully weighing up their options – when it comes to post-purchase satisfaction these people are actually the least happy with their choice!
This was a strong session and an excellent showcase for the application of new methodologies. As chair Vinay Ahuja from P&G said, it showed how listening to real insights from real people is good for your business. But the last word should go to a participant from the Ladies Home Journal community, unwittingly summing up why not only MROCs, but cutting edge research in general is so fun: “never knowing what’s coming next is what keeps me coming back.”
Celebrating Society: Opinion Polls and Democracy. - Anna Peters
Polling is the most accurate assessment of what a population really believes and thinks, allowing for an unfiltered understanding of the temperature of people’s opinions. In this session we came to understand the importance of political polling from three perspectives: we discovered how the ‘simple’ act of polling inspired radical change on the island of Fiji, we learnt how a lack of polling resulted in some great surprises in the Middle East, and we heard that there is an increasing need to protect the practice of polling in the USA.
First up was speaker Caz Tebbutt (Tebbutt Research) who introduced us to Fiji: the land of hot women, strong men and military coups. There have been four military coups in recent years. The first was in the early 90s and it was during this time that Caz established the ‘Tebbutt Times poll’ – an impartial temperature check to see how the nation’s people perceived their leaders. The Tebbutt Times poll lasted 18 years and gave the only view of what the people really thought. This was not always something that politicians wanted to hear, and due to increased military control of the media, 2008 was the last time the poll was run. Suddenly the people of Fiji no longer had a voice.
By way of a substitute, the Lowy Institute launched a poll in 2011. It questioned the people of Fiji on the views on human rights, democracy, and current government. Reaction to the poll was immediate, strong, and widespread: Tebbutt saw over 100 items written about the poll in mainstream media in one week. Of course, media argued over the results and how the poll should be interpreted (and I’m sure that there will be many debates at the hotel bar this evening regarding the ‘right’ conclusion) but as a result of the media frenzy there will now be an election in 2013. This illustrates the importance of polling: it shows polling as a way to give a voice to people who don’t have one, and it shows how polling can have the power to make fundamental changes to society.
Also arguing in support of polling was Nebil Belaam (EMRHOD, Tunisa) who used the “most significant world event in the 21st century” – The Arab Spring – as case for the importance of good polling methodologies in predicting accurate election results. The Arab Spring and subsequent events were the result of the actions taken by Mohamed Bouazizi, the man who set himself alight as a comment on the increasing oppression, poverty and unemployment faced by those under the Ben Ali regime. As a result of The Arab Spring, an election was held on October 23rd. Two months before the election, however, political polling was banned, and whilst Election Day itself was seen as “free, fair and transparent” there was an “election hangover” with many people not expecting the political outcomes that were seen. Indeed, a subsequent poll conducted by Nebil and his team found that 28% of Tunisian people didn’t expect the results, and that 45% of Tunisian people felt that ‘some candidates did not deserve to win as many seats as they did.’ All of this combined offers a very compelling case for the necessity of pre-election polling as a temperature check.
One country very familiar with pre-election poll is the United States: since The Chicago Record’s 1896 poll, America has been polling, and so a very sophisticated industry has developed in this market. Sarah L Dutton (CBS) took us through a brief history of the political poll in American and also walked us through some of the challenges that face the industry today.
She described how (just as in the undemocratic states of Fiji and Tunisia) the main critics of the polls are the political parties in question: if an unfavorable result is found then undoubtedly the candidates and their parties are quick to dismiss the polls as inaccurate, biased or unmethodologically sound. But additional pressures for the polling industry in the States include the language barrier (there is a need to make sure polls are taking into account the increasing Latino population), the rise of technology (pollsters need to adapt their methodologies to take into account these new developments in the way that we interact) and, perhaps the biggest challenge of them all, declining response rates (we have seen a response rates drop by 27% since 1997). She argued that the future of polling was reliant upon good methodologies, transparency, and everyone in the industry doing their bit to defend against criticisms.
My view: “Polls are the pulse of democracy” and they unquestionably measure the voice of the people. The speakers today acknowledged some of the many pitfalls associated with polling, but together they made a very good case for the importance of the poll in democracy. In particular, the case of Fiji highlighted that polling is not an exercise without impact: for the people of Fiji polling really did make a huge change to the way that their country will be run. In short, political polling needs to be protected because it can help leaders make real changes for the benefit of society.
Celebrating Research. Behavioural Economics in Action. - Tom Ewing
It’s sometimes said that the research industry these days is losing touch with its core skills – statistical excellence and methodological rigour don’t get the attention they deserve. The late afternoon session in Parallel 1 will have gone some way to addressing these concerns. An examination of behavioural economics – dear to the heart of research hipsters – concealed the red meat of a serious methodological workout which may have left the unwary attendee somewhat dazed.
Ironically, both papers – from Kevin Karty of Affinova and Florian Bauer of Vocatus – concerned themselves with “softening the rationality of research”, as chair Niels Schiellewaert put it. The presenters took on two of the great cornerstones of quant research – STMs in Karty’s case, conjoint in Bauer’s – and subjected them to scouring critique before positing solutions drawn from the behavioural economics playbook.
Karty’s presentation on improving choice models left the traditional STM no real routes of escape. Most of the accuracy in such models, he pointed out, comes from the embedded historical data, not from the actual consumer score. But that is – or should be – the most important part. Unfortunately in traditional monadic tests the consumer score is fickle, to put it mildly. Karty demonstrated how a series of five parallel tests, on cat litter, gave utterly erratic results, with niche brands outperforming known market leaders and data noisier than the Heineken concept club we’d seen earlier.
What to do about this? Karty demonstrated how discrete choice models worked better than monadic tests, but even these didn’t solve all the problems. Enter behavioural economics, with two simple insights that led to a far better fit between test data and the real world.
The first was the insight that there’s a big difference between our choice and purchase – which to buy, and whether to buy, are separate decisions. To model this Karty included a scaled commitment to buy metric in the choice based survey. The second insight is that most choice based work lacks an anchor point – so by including an explicit comparison to the most common currently purchased brand, we can make the test results still more accurate, and put the niche brands back in their place. Simple changes with measurably better results – the kind of methodological insight researchers have always welcomed.
Florian Bauer was up next, pointing out that conjoint as currently practised contradicts two important learnings from behavioural economics. For a start, the elegant linear steps central to conjoint models ignore the mess of real-world valuations, which are not linear. And secondly, conjoint distorts motivation – the trade-off process induces a level of involvement that’s simply unrealistic. In real life, very few people trade off.
How to solve these? Like Karty, Bauer offered solutions which were simple at heart. To get around the problem of linear steps, conjoint should be combined with price sensitivity measures, which would let researchers spot thresholds hidden in linear data. And to escape the motivational distortion, simply assess high and low involvement groups separately. At the end of the day, though, Bauer warned that these are “patches” for the problems behavioural economics exposes, not truly solutions. Bauer also, in the post-presentation discussions, rejected the idea that conjoint ought to reflect the speedy decisions consumers might make in real world situations – participants already speed through conjoint surveys, he said, the idea was to make them think about them harder.
In the context of the modern research presentation – highly visual, emotional and metaphorical – this session came as something of a shock, and for Bauer in particular the information-heavy style did the content few favours. But that wasn’t to dismiss the insights both men presented. As Karty says, behavioural economics is like a buffet – there’s a lot of great stuff, but if you like something you have to go up and get some more yourself. In that spirit, I’d suggest the best way of experiencing the important work presented here is by sitting down and reading the papers, not viewing the presentations.
ESOMAR Session. White Hat vs. Black Hat: Ensuring the Future Growth of Market Research. - Erika Harriford-McLaren
Privacy in a changing world was the underlying theme of this panel session. Opened by Judith Passingham, Congress Programme Chair, the panel session featured Dave McCaughan of McCann World Group, Lenny Murphy, GreenBook, Sjoerd Koornstra, Heineken International, Mike Cooke, Global Panel Management and Reg Baker, Consultant to Market Strategies International and the ESOMAR Professional Standards Committee.
Each panellist gave a brief presentation providing their unique viewpoint on how privacy regulations are pushing the industry to a focus on social science and stats in a time where new entrants are creating pressure to expand the sector to meet growing and widening client needs.
Dave McCaughen noted that first and foremost, privacy is a consumer issue – in particular the erosion of personal privacy. Although this may differ culturally, it is still a consistent concern across the board.
In general, if people think they will get something out of it, they are more willing to share info about themselves. We must also remember that there is a huge difference between love and trust – people adore Facebook but don’t necessarily trust them with their data. So where do we as MR practitioners sit? Do we want to be loved or trusted?
Lenny Murphy then discussed the role of market research in a data abundant world. With new entrants into the market, we must adjust. When companies like Google Surveys, who own the data, come into play – they will fundamentally change the way we work. He also noted that we are moving into a world of asking versus observing, so much so that brands will soon be able to predict behaviours without asking questions, which could indeed be interesting for the research world and how we work.
He predicted that we will soon be one spoke on the big data wheel and that companies like IBM will be the hub. What this says for the evolution of market research as we now know it, I’m not sure. But it does pose an interesting question as to how we should prepare ourselves for this kind of shift if it is to come.
Sjoerd Koornstra of Heineken International gave an excellent client perspective noting that for a client, and for Heineken in particular, what matters is that an agency has a good understanding of consumers, shopper and market dynamics, while offering global coverage and please… not a lot of slides.
Additionally, and most importantly they want a minimum level of quality… which for them means adhere to the ICC/ESOMAR CODE and ISO Code 20252. For Heineken, Koonstra noted that it would be “unspeakable to use an agency that does not adhere to the ICC/ESOMAR Code. In the end we must trust the data.”
Mike Cooke of GfK believes that privacy goes to trust and transparency. “It’s not that we can’t or shouldn’t use the data – no bones about it we will. It’s about how we use the data. We must give as much as we take, because in the past we took and never gave.
All public research must be testable to underscore credibility. He therefore believes that research should be documented, archived and shared – open to full disclosure, as is done with evidence-based papers with ESOMAR – where all data must be made public so that the public may test if need be. A prime example of we take, we give – and an excellent way to build trust in a data rich world.
Reg Baker, as the last presenter, pointed out that that the success of our industry is balanced on public and regulators believing we will indeed protect consumer data. For Reg it was key to point out that it isn’t just about data quality but making sure your data is legally obtained to help maintain trust and quality in our research.
He noted that the biggest problem we, as an industry, have had in the last 10-15 years has been getting people to cooperate. While technological advances such as mobile are interesting, this is not a panacea.
In the end, we must accurately measure populations – but we cannot ignore standards. As Mike Cooke said at the end of the session….”if we want to have a role in society, it must be one based on evidence and cannot just be judgmental. There is a difference between betting on something and telling someone something is going to happen.”
Keynote: Education in the XXI Century: How can research help? - Kathy Joe
José Ignacio Wert, Minister of Education, Culture and Sport of the Kingdom of Spain, discussed the role of research in developing education fit for the 21st century. An ex-researcher himself who specialised in opinion polls and social research, José Wert is intimately aware of how governments and politicians can benefit from the contribution of research.
He noted the importance of relevant education in ensuring that young people are fit for the workplace of the future, saying that the impact of a global economy means that increasingly people will have the chance to rise in their careers based on what they know rather than their social background.
José commented, “Education in the 21st century is the main engine of growth for jobs and equality for every economy and society but it faces huge challenges at all levels and in every society. One of these key challenges is that the formal education process requires a radical transformation of the teaching and learning environment to fully take into account the immense possibilities that technology provides.” He added that research and particularly technology-based research can play a major role in guiding and assisting this transformation.