Download PDF of Article

The Black Ops of Retail

April 17, 2012 

By Teresa Novellino

To discern what consumers want, the retail industry uses samples, in-store tests, and focus groups. But a startup called First Insight says it can predict what consumers want even before the product exists.

In the retail business, they’re called dogs: unsold items that are placed on shelves and never roll over, items that, at best, will sell at a loss. Their similarly troublesome counterparts: popular items that customers snap up to the point that retailers run out, leaving disappointed customers and potential dollars on the table.

Separating the winners from the losers is the job of merchandising executives, who visit stores, hold focus groups, and test-run merchandise in a handful of stores. Some are using a new ally: a Pittsburgh-based software company called First Insight. The company discerns what consumers want by having them—believe it or not—play online games.

Greg Petro founded the company in 2007, but up until the last year or two, he has kept it under the radar, quietly helping retail clients and manufacturers figure out what to buy and what not to, and even what consumers will pay for items that have not yet been manufactured.

“Our whole perspective has been to be the black ops of these companies,” Petro told in a recent interview. “Five years ago, we started creating these games, these little widgets. We’re now collecting a million pieces of data a month from consumers who [might not want to take a retail survey but will play games] all day long.”

Petro, a trained economist and former Saks Inc. executive started his retail career in the 1980s, and as a former merchandiser, he understands the odds well. New products have a failure rate of 50 percent and more, according to data from MIT Sloan Management.

Among First Insight’s games is one that lets users guess the price and rate an array of items. If players are shown images of four different watch models, for example, there might be two that already exist and two that do not. The as-yet-not-produced designs may never see daylight if they don’t score high enough among the consumers surveyed.

“We’re not telling them not to make it, we’re telling them not to make a lot of it,” Petro says.

But not all consumers are equal, and that’s part of First Insight’s secret sauce. Its algorithm gives more credence to consumers who are particularly savvy about prices or who are especially knowledgeable about certain types of products. Yoga devotees who practice every day are separated from weekend warriors, for instance. Consumers in an urban West Coast zip code are distinguished from those in a Midwest suburb: That way the games can yield store-specific information.

Among those openly using First Insight is David’s Bridal. Mass- producing a wedding gown made in pricey fabrics like silk that can cost as much as $1,400 can be a profit-killer for the company, but with First Insight's gamification tools, it has added 20 percent to its top-line sales and shaved three months off a production cycle that for apparel is typically eight months.

The company goes to a China manufacturing facility, takes photographs of gowns, and sends them back to First Insight in real time. Images of those dresses are sent out to consumers in the form of a game. Within three days, the results are in.

“It’s giving them information that they never had before in cycles that are 48 to 72 hours,” says Petro. “By the time they’re back in their offices outside of Philadelphia, it’s all available to them.”

In addition to using real dresses that have been created, First Insight can test computer-designed sketches that came straight from a designer’s imagination but do not yet exist, saving on the costs of expensive samples.

Retailers in First Insight’s portfolio have seen comparable store increases of 3 to 9 percent since partnering up, and gross margin increases of 4 to 9 percent. The games might appear on a retailer’s website, or they might arrive to consumers via email, social media, or whatever tool the retailer prefers to use.

Petro says that he kept quiet about the company at first because he didn’t want to alert would-be competitors.

“We didn’t go public until 18 months ago. [Most people didn't know] that we even existed,” he said. “We were working with retailers, and we didn’t need to [go public]. We wanted to make sure that we had a competitive foothold and a number of data points [so] that we can stand on our own two feet. [Otherwise] Oracle and IBM could walk in and say that they’re going to do it. Now we’re two years ahead of everyone.”

The company works with close to 20 retailers, most of which do business in the $10 billion annual revenue range per year, though the span stretches from retailers who do $50 million to as high as $70 billion. Together, its clients represent a total of $190 billion in sales a year out of a U.S. market of $2 trillion.

Because First Insight remains in a “high-growth phase,” it doesn’t discuss revenues or the size of the company. Amid the talk of social shopping and big data, it seems that technology like this is very much going to be a part of retail’s future.

Petro says his company’s efforts are different from others in that it is putting the consumer at the wheel in the very earliest stages of merchandising.

“We’re actually democratizing and involving the consumer in the decisionmaking,” Petro says. “What we’re doing and what we have done is fundamentally changing the retail mantra of how companies operate. And no one in the world is doing this outside of us.”