At a recent ARF conference, Stan Sthanunathan of Coca-Cola exhorted the market research industry to move beyond understanding consumer needs to understanding consumer motivations. If we are to accomplish this, we need to go beyond observed behaviors and their attendant inferences to truly immerse ourselves in the “why”: why consumers choose one brand over another; why they decide to “like” something on Facebook; why they buy certain products at certain stores.

Our questions must be carefully framed if we are to arrive at appropriate answers. So the question “What drives people to ‘like’ my brand?” is more likely to lead to valuable insight than is “How do we increase the number of people who ‘like’ our brand?”

In his seminal article titled “Risks and Riddles” in the June 2007 issue of Smithsonian magazine, Gregory Treverton points out that questions can be thought of as either puzzles or mysteries.       

When we solve a puzzle, we end up with a specific answer — one that is quantifiable, comparable to answers to other puzzles treated in a similar manner, and invariably a go/no-go decision. That’s how business works We solve puzzles, make decisions, and move on.

By contrast, mysteries, even when “solved,” don’t necessarily result in quantifiable answers that dictate a business direction. To quote Treverton, “Mysteries pose questions that have no definitive answers because their solution depends on a future interaction of many factors, known and unknown. A mystery really can’t be answered; it can only be framed by identification of factors that interacted in the past and may in the future.” Diagnostic approaches to problem solving invariably follow the mystery paradigm. We start with an issue — declining sales, for example — and attempt to explain it by examining different scenarios, much like a physician ordering test after test to uncover the cause of a malady.

Solving a mystery is far more rewarding than solving a puzzle — in terms of both financial returns and personal satisfaction. But mysteries in business are a bit like emotions in the boardroom: uncomfortable and to be avoided wherever possible. And if we have been especially successful at solving puzzles, we may be tempted to define all business problems as such (because “If all you have is a hammer, everything begins to look like a nail.”)Millward Brown: Point of View Solving Puzzles Delivers Answers; Solving Mysteries Delivers Insights         2

A puzzle analysis tends to be somewhat linear and focused on a specific goal; we know when we have reached the end. But the path to unravel a mystery is anything but linear, and one question tends to lead to another. While puzzle solving yields answers, resolving mysteries yields insights. We cannot be entirely sure what we will find when we undertake to investigate a mystery, and thus some degree of hypothesis development or scenario building is appropriate to define the acceptable level of ambiguity in the outcome.

Market researchers address both puzzles and mysteries. Our puzzle questions include: “Which household member makes the brand choice in a category?” “What factors influence those choices on different purchase occasions?” “Which of three TV ads should be aired?” “Did an ad campaign produce the intended increase in awareness, trial, or sales?” “What media mix is most effective? What is the risk in increasing prices by 5 percent?”

Our mystery questions include: “What will the opportunity be for a new type of vacuum cleaner?” “Will emergent markets such as Brazil or South Africa gravitate to global brands or to locally developed ones?” “Will consumers accept a totally revised brand formulation?” Coca-Cola learned a hard lesson related to the last point. The launch of New Coke — the greatest marketing blunder since the Edsel — took place because Coke’s decision-makers mistook a mystery for a puzzle. Yes, the taste of New Coke was preferred in blind taste tests, but the real question should have been “How will consumers feel about a new version of a beloved and iconic brand?”

To penetrate a mystery, we often need to make a paradigm shift; we need to think about the world in a new way. Henry Ford acknowledged as much when he said, “If I’d asked people what they wanted, they’d have said a faster horse.” However, what Ford probably addressed was a mystery question: “What does the future of transportation look like?”

The Challenge of Too Much Data

With the introduction of the Model T, Henry Ford ushered in a new era of transportation. Likewise the advent of computers, scanning technology, wireless communication, and social media has led us into a new era of market research, one in which traditional methods of “active” data collection are now supplemented by a profusion of passively collected data. We can now track website behavior, clicks, likes, tweets, and blog chatter. We can download reams of scanning data and information from shopper loyalty programs. An enormous amount of such passively collected data is generated every day.

This data is invaluable for answering the many “puzzle” questions that start with “who,” “what,” “where,” “when,” and “how.” The answers to these questions tell us what people are doing now — but not why. We need motivational research to help us fully understand the answers to questions that start with “why.”

Some would say that “Why?” is not an important question anymore. On Research in March (2011), Jules Berry noted that behavioral economics has undermined the notion that our actions can be predicted at all through direct questioning. He reminded us that many analysts recommend that we skip questioning altogether and go directly to observations of actual behavior, sensory experiences, and brain activity. Like members of the behaviorist school of psychology, many data analysts seem to believe that as long as we can manipulate the environment to perpetuate or grow desirable behaviors, we needn’t bother asking “Why?” So we experiment and manipulate stimuli, each time seeking to

improve response rates and ultimately revenues and profits.

New industries and subspecialties of existing disciplines have emerged to help companies stay afloat in this ocean of data. For example, IBM has invested more than $10 billion in companies dealing with predictive marketing analytics. Tesco spends more on mining its shopper loyalty card database than on all other forms of consumer and shopper research. Reacting to the “findings” in this sea of information requires measurement, norms, fast decision-making, and further experimentation.

While this information can be valuable, there is a real risk that most marketers will attempt to make sense of it all by treating it as “grist for the puzzle mill.” And if we make marketing and business decisions exclusively on puzzle solutions, we impede the advancement of innovation. Every new idea is going to look a lot like the old ones — only a bit slicker. The risk is that if marketing moves exclusively to iterative incrementalism, we will limit the potential for breakthrough ideas that are based on a deep understanding of human behavior.

What mystery questions require is not more data but better analysts.

Would we have a brand as successful as Apple if all Steve Jobs sought was a faster and more reliable computer? His insight was that people were open to using technology to connect with others, to have fun, to get information — in short, to enhance their lives, to be empowered. What rate of click-through data would have delivered that? And what about Starbucks? Coffee was becoming purely commoditized when Schultz understood that it was much more than a beverage. Taking his lead from the culture of European coffeehouses, he brought “the third place” mindset to America — and subsequently back to the rest of the world. And what of Method, or Body Shop, or Red Bull? Each was founded on a deep understanding of human motivation and is being maintained not by churning out line extensions to gain coveted shelf space, but by careful consideration of what is right for the specific audience for the brand. Each has a good grasp of the motivations of its customers.

So What About All That Data?

While analytic approaches to puzzles are different from those of mysteries, both can draw on passively and actively generated data. However, puzzle questions often require more data, and in fact, the more data applied, the better the solution is likely to be — up to a point. But, as the vast sums spent by IBM and Tesco indicate, the management of this data carries a hefty price. At some point, companies could find themselves in a “Mutually Assured Destruction” phase of data gathering and analytics, where terabytes and processing speed become more important than understanding what it all means.

By contrast, what mystery questions require is not more data but better analysts. Mystery problems, when cracked, can lead to competitive advantage, whereas puzzles solved lead to confirmatory judgments. Particularly when so much data is available to everyone, there is a limit to the competitive advantage to be found in data mining. Hence the need for unique information and insights. As pointed out by Thomas Friedman in The World Is Flat, what we do well, so too do our competitors.

So What Does This Mean for the Researcher?

We should look at all problems as both puzzles and mysteries. If we only frame an issue as a puzzle, we will get an answer. However, by also framing the issue as a mystery, we open the solution to far more options. So instead of just asking the question “How many units is my advertising selling?” we should also be thinking in terms of “Is advertising the best way to deepen the relationship people have with my brand?” This opens us to possibilities that may lead to far better approaches than we currently have.

Puzzle solving and mystery solving require different mindsets and potentially quite different skills — perhaps different people entirely. Invariably we will need to hire and train for both skill sets. Most market researchers are trained to solve puzzles and, for the most part, the suites of software developed are

designed to arrive at convergence; i.e., the answers to puzzle questions such as “Are these responses significantly different from those?” “What is the ROI of the advertising?” “Which group of people is most likely to buy the brand?”

By contrast, the industry hasn’t given sufficient attention to mystery-solving skills. There are data explanatory tools, but as long as the focus of marketers is on analyzing the tidal wave of behavioral data (nails), hammer skills will be in demand. Mystery-solving require less data but greater imagination — the ability to triangulate information and exercise intuition. Therefore we should be training mystery solvers for divergent rather than convergent thinking and for creative problem solving rather than statistical procedures. We should be competing with intelligence services for the right kind of analytic minds.

Finally and most obviously, we need to respect and appreciate both puzzles and mysteries, and use the most appropriate resources to answer the type of question we are facing. Data by the bucketful can help answer questions that have definitive answers and lead to sound business actions, while smart analysts can tackle intractable problems as mysteries. Invariably most business problems are puzzles, but reframing them as mysteries can generate deeper insights and potentially far greater rewards. By embracing the pursuit of mystery and remaining firm in our belief in the irreplaceable value of motivational research, we can meet the challenge set out by Stan Sthanunathan. But if we answer only the questions our clients voice — if we settle for just solving the puzzles — we risk a future of well-defined mediocrity.