In the coffee-bars around the Market Research Society’s annual conference this year, the talk was all about jigsaw puzzles. Before a deluge of digital data descended on their heads, brand owners turned instinctively to market research to learn about customers. Now market research is just one of umpteen routes for discovering things about consumers – hence, the preoccupation with jigsaws.
Analogies are useful. But, beyond a certain point, most run out of steam. In the puzzle solver’s world, the pieces of a jigsaw slot snugly together. In business, data fragments take myriad forms. And whereas the building blocks of a jigsaw come handily packaged in a box, the data needed to make sense of customers is more likely to be scattered across a labyrinth of databases than housed, conveniently, under one roof. So, do market researchers, steeped in their own craft-based niches, have it in them to become masters of jigsaw-assembly?Mike Page, a former operations director with Synovate, has doubts. “If you ask an agency to integrate all your data and some of that information comes from a competitor agency, there’s an immediate tension. Research suppliers worry a lot about revealing branded techniques that a rival might replicate,” he says.
To occupy what he saw as a gap in the market, Page left the market research world to set up Cognicient, an agency specialising solely in insight integration. Working with a combination of software-based analysis and human scrutiny, Cognicient culls fragments of data from multiple sources, and puts them into a format capable of building a coherent picture.
“The market research industry is set up to spit out answers in response to a brief,” says Page. “But, the answers aren’t structured so they can be combined with other sources. The opportunity for us is to turn all those one-off studies into jigsaw puzzle pieces that can be fitted together.”
Multiple data streams
Market research agencies, not surprisingly, contest Page’s view that they are fixated on the short-term and incapable of working co-operatively. But the investment brand-owners are making in customer databases leaves few agencies in any doubt that the market research industry must grasp the nettle of working with multiple information feeds.
One possibility is for market research agencies to partner with experts in data handling. As the data specialists behind Tesco’s Clubcard, for example, Dunnhumby knows all about people’s buying habits. But what its transaction-based datasets have historically lacked, and what market research can add, is the ability to explore the hidden stories that underlie observed purchasing patterns.
To take a simple example, a customer of a supermarket might have purchasing habits that suggest a concern for healthy eating, yet never buy organic produce. One possibility is that she thinks organic is no healthier than other produce. Another explanation is that she buys organic, but prefers specialist suppliers, such as Whole Foods Market or Abel & Cole, to the organic counter in stores. Both scenarios are plausible, based on the data held on the retailer’s systems. But, only by probing behind the bald facts, through market research, would the store owner find out which explanation is more likely to be correct.
“If we are trying to understand a customer’s purchasing history in order to improve the quality of our marketing, we can only do that by combining transactional data with qualitative research,” says David Inglis, group trading director at Shop Direct Group, which has recently begun working with Dunnhumby to segment its customer database and improve the accuracy of its targeting.
One way in which market research scores over consumer insight derived from purchasing records is in its ability to tell brand-owners and retailers about the sales that they lose out on to competitors. In the short term, this need to plug gaps in the competitive picture creates an opportunity for market researchers to sell tracking surveys to brands. But there is a penalty to pay for relying on customer hearsay. Ask people what they buy and they may tell you, but they may lie or simply not remember.
Sources of intelligence
From a strategic perspective, then, the opportunity for market research agencies is not to position themselves as suppliers of make-weight proxy data. Rather, it is to capitalise on the explanatory power of market research and to seek to harness that strength to other sources of intelligence. Here two approaches are possible. One attacks cultural barriers standing in the way of data fusion; the other tackles technological hurdles.
From the qualitative side of the profession, Jamie Allsopp, head of innovation development at Sparkler, detects a space opening up for agencies that help brands to decide which trends to follow and which to ignore. As part of its offering, Sparkler will trawl through a client’s research archive; distil the key trends and facilitate workshops in which managers come together, from across the business, to devise a strategic response to things that are happening in the market. “Companies have research coming out of their ears; but often it’s held in departmental silos, or locked up in people’s heads,” says Allsopp.
At the other end of the spectrum, TNS is dabbling in digital surveillance. A long-established provider of consumer surveys that rely on participants dutifully logging what they buy ©on household scanning equipment or making entries in dairies, TNS recently purchased a US business, which has recruited 2 million consumers willing to host click-stream software on their PCs.
With the new technology, TNS sees every purchase that panel members make online and every step taken en route to the transaction. John Abraham, head of strategy and marketing at TNS Global Business, takes, as an example, a consumer who buys a pair of Nike trainers: “From what someone looked at on the Net, we know which brands they considered. From the product features they clicked on, we know what parameters were important to them. For example, was it shock absorption, colour or price?”
As more data on consumers’ buying patterns becomes available, a risk for market research is that brand-owners may decide to run marketing campaigns based solely on their understanding of what they observe customers to be doing, without concerning themselves overly about why they are doing it. In the long-term, this might prove a false economy.
“From someone’s browsing behaviour, we can deduce that they are concerned about shock absorption. But, to market to them effectively, we need to fine-tune what we say,” says Abraham. “By exploring why particular groups think shock absorption is important, we can begin to serve up advertising that is targeted at different profiles of runners.”
Aiming different messages at different customers is already happening on the internet. In an ideal world, to target someone with perfect confidence, a researcher would study an individual’s buying habits, then interview them personally. In reality, marketers aim products at individuals based on statistical predictions of what might interest them, derived from what is already known about their tastes. Some predictions prove correct; others wrong.
“Although data fusion offers unprecedented insight into consumers as groups, there is a danger that once you try to target individuals, you begin to assume things about them that aren’t true,” cautions Andrew Corrall director of analytics at customer reward specialists LMG. “A person’s attitudes are very hard to predict, because, ultimately, there is a limit to how much we can know.”
One issue likely to loom large in the market research industry’s future is consumer privacy. As chairman of the Market Research Society, Rowland Lloyd, managing director of operations at Ipsos MORI, makes it clear that agencies working with consumer data owe a duty of care to the public. “If data is identifiable to an individual, we need to be careful that consent has been given. Equally, as more and more data is shifted around, there needs to be a proper investment in the encryption and secure movement of files to ensure that we don’t end up with data getting lost.”
How the public will react to personal advertising is, as yet, unclear. Along with the rest of the marketing industry, market researchers argue that behavioural targeting, based on people’s personal tastes and habits, is a good thing. For Lloyd, the only worry is that the execution may not live up to the promise, so consumers could be deprived of advertising they might have found useful.
“If it is done badly, people may miss out on information that they would otherwise have been made aware of. But on balance I’m pro it, because being sent stuff that’s relevant to me is better than being sent stuff that isn’t.”
Now, consider a different scenario; one in which you are constantly assailed by a stream of advertisements, all perfectly attuned to your tastes and habits. Would that be marketing heaven or consumer hell? Marketers may wish for perfect knowledge of consumers; but if perfect knowledge were possible, as consumers of other brands, they might yet rue the day.