Whether they call it a gut feeling or instinct, most marketers and researchers say they know a good insight when they see one. Beyond that, there isn’t all that much agreement.
The only good insight is one that solves a business need, says one camp; it’s whatever resonates with shoppers, says another; bottom line, it should help increase sales, says another.
There is agreement, however, on one critical point: Today, a consumer/shopper insights team must rely on a larger, more formalized process to uncover and validate insights. Gathering the research, mining the data and translating it into insights and actionable creative requires a skill set that’s dramatically different from just five years ago, says Craig Geiger, director, category development and shopper insights, Barilla America. Back then, he says, a data miner could simply pull and organize data; today, it’s that plus 50 pages of research on how one insight can be applied across multiple teams.
“There’s clearly the skill set from a research standpoint, and that’s the ante: everyone needs to deliver strong skills there,” says Deb Fifles, vice president, consumer insights, Safeway. “But as important, we look for folks who have strategic broad thinking or are able to tell a story – how do you connect your business teams with that insight? It’s the folks who can sell what they do in a credible, meaningful way.”
So what constitutes a good insight? Fifles looks for one that carries a lot of shopper emotion, be it positive or negative. “When we’re really looking for an insight to leverage for a communication campaign, we’ll look for the things that strike a chord with shoppers. Then we’ll look for data to support that – quantify after the fact.”
For example, in 2010 Safeway completed a week-by-week study that looked at how shoppers behaved leading up to the holidays. The information influenced communication for the 2011 holiday season. The study covered what activities shoppers did with families and friends, and how they felt about those activities. “They took pictures for us. They talked about where their stress points were, what made them feel good during that timeframe,” Fifles says. “Everyone will tell you they’re busy, and they’ve got a lot to do, and they go to a lot of parties, and they make a lot of food. Most folks have a turkey on the table for Thanksgiving, but understanding more around that is where you have the opportunity to differentiate.”
Geiger says he looks for the “So what?” of the data. “Does the insight address the issue or opportunity? Does it shed light on our hypothesis?” He gives an example of launching Barilla Piccolini, miniature pasta that comes in fun shapes such as a star. Geiger’s hypothesis was to position all of the Piccolini shapes for kids, including babies as they transition into solid foods. But retailer shopper card data unveiled that only certain shapes worked for that while others worked primarily for soups.
Christopher Farley, director of insights, North America, GlaxoSmithKline, follows three criteria to determine a good insight: Is it “Real,” or does it reflect a deep understanding of behaviors or attitudes? Is it “Relevant,” or does it apply to the business? And does it “Resonate,” or do shoppers relate to the insight? “I don’t think this 3R model is foolproof,” he says, “but I do think it provides a solid framework from which we can whittle out bad insights.” He adds that he no longer tries to listen to the debate of what defines a good insight. “Many business leaders will suggest that insights can only be deemed ‘good’ when they lead to business success, but I don’t agree with that perspective, because often solid insights are executed poorly.”
Michael Ross, vice president of marketing, pricing and consumer insights, at Meijer disagrees slightly, saying good insight equals business adoption and action, and a sales increase. “An insight accomplishes many things; most of them are strategic yet passive in nature. Cost aversion, improvement and identifying drivers seem to be at the top of the list on what insights tend to solve.”
Leaning on more of a feeling, Kathleen Wolf, senior manager, consumer strategy and insights, Whirlpool Corp., is looking for that “aha” moment. “An insight is stimulating information that typically comes from an understanding of the raw data in context of a bigger picture,” she says. “It’s a fresh, not-yet-observed understanding that creates pause.”
While shopper marketing and insights will always be a mix of art and science, spotting a good insight seems to be more instinct and business sense than anything. “Because I believe so strongly in the fact that a good insight is rooted in understanding the bigger picture, I would say that it’s more instinct vs. science and process,” says Wolf. “Information is science. The art is creating insights out of the information. It’s the smashing together of information that creates the most actionable insights.”
Fifles says it “feels less science to me than instinct. You can get good insights from anywhere. It can be an observation when you’re doing something that has nothing to do with grocery shopping but it triggers a thought, an analogy and there’s an aha moment. Then you build your case around that, and that’s where the system and process comes in.”
Scott Young, president of Perception Research Services, agrees that the art is in translating an insight into a testable scenario and action. "We often gather insights about shopping behavior from our in-store studies. The challenge is then applying or translating that insight into new scenarios or actions - such as new packaging systems, aisle configurations or in-store merchandising - that we can assess among hundreds of shoppers, to see if they make a difference."
Meijer’s Ross says good instinct drives good insights. “Analytics should drive two-thirds of the insight. The rest should be shaped by business sense and actionability. Shopper marketing is science and art, not one vs. the other.”
Matt Kleinschmit, senior vice president at Vision Critical, a market research company, says instinct is really code for experience. For them it’s talking to clients and understanding CPG and retail to know what will succeed, but “we do believe that even good insights can’t be vetted without strong research science behind them. Otherwise the insight is simply a hypothesis with potential.”
If instinct helps identify the insight, then science qualifies it. As Ross says, “triangulation of results is king. A number is not thought of as true unless it can be validated or confirmed with other sources or within additional context.” Ross also says sample size is important to ensure broad representation of the population, but there needs to be a consistency in answer patterns.
Young validates an insight through quantitative behavioral testing of new retail scenarios. For example, through in-store research, PRS may observe that shoppers' behavior in a particular aisle suggests a potential opportunity. Then, they'll work with a client to develop and test a new retail scenario, based on that insight, among hundreds of shoppers. If the new approach drives growth (of the brand and/or the category), they'll know that the insight - and related action - has been validated.
Barilla’s Geiger has a system in place for mining and sharing data that he calls “elegantly simple.” What’s simple is the business is basically pasta and sauce. “Guess what, Americans dump sauce on pasta and eat it like a pasta meal. But I want to peel it back — what’s the pre-store, in-store, post-store decision? Why did they select the dollar store versus mass versus grocery when buying premium sauce? Are they stocking up on spaghetti because the husband likes it or because the price is right?”
Geiger says one of the strengths at Barilla is how it formed around insights but clearly defined the usage of insights by functional area. “And not to be trite, but I have really boiled it down to: MAPS, TILE, TAPE and BUDS”:
Barilla starts with projects, initiatives or hypotheses to review, and then each function assigns team members to source data in three areas: custom research, syndicated data and customer data. Barilla also leverages someone from Nielsen and someone from Advantage Sales and Marketing. They package the insights for the teams, and the insights are available to everyone, as needed. Upon reviewing, the teams also determine gaps in knowledge and if more research is needed. “We can tell in a heartbeat if we’re light on post-store understanding of how people are using that pasta once they buy it,” Geiger says.
With so much time, effort and money put into an insight development initiative, there is an urge to settle on an insight to justify the work, but the key is to resist the urge and keep mining, according to GlaxoSmithKline’s Farley.
Fifles admits that sometimes her Safeway team may feel like it’s settling because it has run out of time and knows it can do more, but she says the team continues to work so it never really settles.
Perception Research Services’ Young believes there are good insights in every study. The problem is there can be a tendency to bury the true insights. “There’s always the temptation to report everything because that’s what people paid for. The challenge is always to distill it and to say here’s what was really interesting.”
Meijer’s Ross says the goal isn’t to create an insight or settle on an insight. “Classic misunderstanding about the role of insights,” he says. “The goal is to enhance a business process or service.” And more important than the cost of insights themselves is what can be done with the results, he says.
Says Whirlpool’s Wolf: “Gathering information justifies the cost and expense. Insights are the gems that are rare and unexpected. If every project created an insight, I don’t believe we are using the same definition.”