Ideo Spinoff ShopWell Says Better Health Starts at the Supermarket; Part 3: Food as Data
These days, the simple act of going to the grocery store is a fraught and anxious affair. Americans are being told that what they choose to eat isn’t just a personal decision, but has major economic, political, and moral implications. For one thing, there’s the spiraling cost to society of food-related health conditions, from obesity and diabetes to heart disease and hypertension. Then there are books like The Omnivore’s Dilemma and Eating Animals and movies like Supersize Me, Food Inc, and Our Daily Bread, which expose the unsavory sides of a food economy dominated by supermarkets and fast-food joints and the factory farming system that’s grown up to serve them. And, of course, there are the burgeoning organic and local food movements, which argue that foods produced locally and without the use of pesticides or antibiotics are healthier and more sustainable—even if they’re beyond many consumers’ price range.
Into the middle of all this steps ShopWell, a Silicon Valley Web and mobile startup spun off last year by the design consultancy Ideo. The company’s proposition to consumers is simple: tell us a little about you, and we’ll tell you which products on the supermarket shelves best fit your nutritional needs.
It sounds great, given the difficulties ordinary mortals face with the very first step of responsible shopping and eating: trying to figure out how ingredient lists and nutrition labels relate to their own lives. But the company’s plan for making money is a bit more complex. It wants to be an advisor and information broker to food producers, who supposedly lack good data about how consumers make buying decisions in the grocery store and therefore have a terrible record when it comes to launching new products. To collect useful intelligence for its food-industry clients, ShopWell will need lots of users. And to sign up lots of users, it will have to provide non-obvious product recommendations in a usable format.
But frankly, it’s not there yet—which isn’t surprising, given that the company launched its beta site in September and its iPhone app even more recently. So in this third and final installment in our ShopWell case study, we’ll look at the product development challenges the company has ahead of it, and the business-model hypotheses it has yet to test.
What makes ShopWell worth following, and sets it apart from the scores of Silicon Valley startups launched every month, is not just that it’s trying to validate two premises at once (i.e., that consumers want an easier way to identify healthy food, and that food producers will pay for fine-grained data about consumer preferences). It’s also that the startup’ fate will reflect on Ideo’s ability to launch successful companies. Is the consultancy’s fabled user-centered design philosophy an effective tool in the real rough-and-tumble of startup life? Do good designers also make good entrepreneurs? Such questions may not be answered until ShopWell itself exits the startup market and its venture cashiers ring up the totals.
The Google of Food?
Whatever else it may mean, the “user-centered design” philosophy espoused at Ideo and many other creative hotbeds is about listening to people and creating things they’ll want, rather than force-feeding them products that don’t fit with their existing behaviors. But one of the interesting things about food and wellness, according to ShopWell CEO Jasmine Kim, is that many people only start thinking about the subject once they’re forced to give up their old behaviors. “A big insight from user-centered design”—that is, from the stories people have told Ideo and ShopWell—”is that when you are starting a transition, that is when you need nutrition advice,” Kim says. “You could be told by your doctor, ‘You have type 2 diabetes, eat less sugar,’ but people are left to their own devices to go figure out what they could eat.”
That’s why ShopWell has put a lot of energy into allowing users to filter its database of foods based on health needs—there are simple “preselects” on each users’ account page that will adjust ratings downward for whole categories of foods that might aggravate a medical condition like diabetes. And that’s why it’s trying to build the Web’s most complete and up-to-date catalog of foods. Eventually, the catalog will include everything regulated by the USDA (fish, meat, poultry, fruit, and vegetables) and almost every food product with a UPC barcode, including store brands. (But it could take a while to get there—SymphonyIRI, a consumer packaged goods consulting firm, tracks at least 10 million separate stock-keeping units or SKUs in the grocery sector alone, according to Kim.)
But there’s another reason why people negotiating transitions in their eating habits are important to ShopWell: that’s when they’re actually thinking about food choices, and are therefore in a better position to educate food producers about what’s important to them. “I think the reason we have so much interest from potential customers is because they know that when people are this committed about nutrition and health, that is one of the most compelling things that can move the marketplace,” says Kim.
If major food companies are bad at predicting what consumers will buy in normal times, they’re doubly unprepared for a general shift toward healthier eating—an era when to use Kim’s words, “even people at Wal-Mart are demanding some level of organic” in the produce aisle. This fact has everyone at ShopWell, at its mother ship Ideo, and at its venture backer, New Venture Partners, convinced that the data the startup collects about consumer preferences will be highly valuable to food marketers.
“The real value of ShopWell is that they have figured out a way to monetize consumer traffic that is an order of magnitude more effective than selling clicks or impressions, and that is on the quantitative analysis and insights side for food producers,” says Robert Rosenberg, one of the partners at New Venture Partners who helped engineer ShopWell’s spinout and arranged a $2.4 million seed investment.
The genius of ShopWell’s system, Rosenberg says, is that in order to get personalized food ratings, ShopWell users “voluntarily part with personal information of the sort that marketers would kill for, figuratively.” And the startup has a channel for gleaning more and more personal information over time, about users’ health conditions, gender, age, and activity levels. It can even glean rough geographic information from users’ IP addresses, as well as the local brands they’re putting on their shopping lists. “So when consumers use the site, they leave this trail of breadcrumbs as a side effect, and this is where the insights for food producers come from.”
All of the data ShopWell plans to provide to food companies would be anonymized, of course. But it would theoretically be far more fine-grained than anything the industry is able to get by running a few small focus groups.
Say, to use a completely fictional example, that Unilever is trying to figure out why Kraft’s Miracle Whip is suddenly outselling Unilever’s own Hellman’s Mayonnaise in the deep South. They might learn from the products that ShopWell users are putting on their shopping lists that Southern women with heart disease are three times more likely to choose Miracle Whip, which has less fat and cholesterol. That’s information Unilever might be able to use to create a heart-friendly version of Hellman’s, and then market it in specific regions. “When you have the world’s largest focus group telling you what they are doing, you can start slicing and dicing and segmenting on a Web scale,” says Rosenberg.
Kim says ShopWell will also offer food marketers direct access to its users, presuming they’ve indicated that they’re willing to be recruited for surveys. “If you are a marketer, you could test all aspects of product pricing and promotion, by recruiting a panel of people who have said they’re willing to speak to manufacturers. That’s smaller, but it’s real-time, and it gives you enough data to make quick decisions in an industry that is very risk-averse.”
Kim calls the ShopWell market research model “Google Analytics for the food industry.” But that comparison may actually undervalue the service, since Google Analytics is free, and food producers are used to paying millions of dollars per study for market research.
“Anything that costs less than that will be great,” Kim says. “And we’re not only giving them the data, but a platform for dialogue, so they can go back to the people who gave the data and say, ‘Hey, what if I did this, what would you think?’”
This entire vision, however, hinges on signing up enough users to yield actionable data. That doesn’t necessarily mean ShopWell has to grow to Google scale; “Because of their hyper-effective monetization model, they don’t have to get very many million to be wildly successful,” Rosenberg asserts. But at a minimum, ShopWell will need enough users in each major geography and demographic group to yield statistically valid insights into trends in the larger population.
To gain those users, ShopWell will probably need to improve on its existing service. While the ratings are interesting and the food catalog seems exhaustive, ShopWell doesn’t yet include many tools to make the data useful in contexts outside the site. It would make perfect sense, for example, to be able to access your Web-based ShopWell shopping list from the ShopWell iPhone app, or at least to export your ShopWell shopping list to common task-list management tools such as Remember The Milk. But so far you can’t even e-mail your shopping list to yourself or your spouse. Nor is there an application programming interface that would allow outside developers to build tools that integrate with ShopWell and fill such holes.
To get the site launched, Witlin says, his team followed the “minimum viable product” precept; it’s a part of Eric Ries’s lean startup philosophy that says developers should only put enough features into their product to start eliciting user feedback. In ShopWell’s case, the minimalism shows. But if the profusion of Post-it Notes around the ShopWell office is any indication, there are a lot more features on the way. And in theory, ShopWell ought to have one big advantage over other lean startups: with its Ideo inheritance, it should be able to avoid random iteration and instead focus developers’ time on features that have been pre-qualified by potential users through ethnographic studies.
If ShopWell can ultimately help consumers and producers cut through the current maze of conflicting, shrill, and often emotionally and politically laden messages about food, it will be doing a service. Yes, food producers are mainly out to make money—but they also recognize, Kim and Witlin say, that today’s consumers are increasingly concerned about health and nutrition, and they know they’re going to have to confront the issue somehow.
“We believe that anybody can change, whether it’s the consumer making a tradeoff to eat better for their family, or a food producer reacting to consumer needs by reducing sodium levels or pumping in less high-fructose corn syrup,” says Witlin. “By sitting in this spot and not being judgmental and creating this increased transparency, we think we can move the needle.”