Not long ago, specialty retailers like Crate and Barrel depended entirely on people who worked in their brick-and-mortar stores to connect with customers. The in-store sales associate—let’s call her Michelle—was and remains someone who knows what her customers are looking for and are most likely to buy. Because after years of working with customers in one of the chain’s ninety-two outlets, she has an instinctive feel for what is on the mind of the shopper.
But with the explosive growth of online shopping, there needs to be a digital Michelle to help guide and inspire the shopper. That’s the goal of SLI Systems, an e-commerce accelerator that partners with Crate and Barrel to anticipate what digital shoppers might be looking for, and then insure they can find it. As SLI vice president of sales Carter Perez explains, the company’s e-commerce solutions basically crowdsource information from online shoppers to expedite and magnify the shopping experience—and to maximize profitability to the retailer.
That’s more than digital marketing-speak. In the second month of its program with the iconic housewares and furniture retailer, the conversion rate (site visitors who conducted a search and purchased something) went up 5.4 percent. For a mature online retailer, that’s no small dish. “That kind of growth is hard to achieve in a traditional store,” Perez says.
SLI employs “machine learning” to accomplish this. A subset of artificial intelligence, machine learning evolves product ranking according to an algorithm that looks at site visitor search terms. With each subsequent search and visitor, the ranking is fine-tuned. In other words, as the crowd (cumulative site visitors) gets bigger or on-site activity increases, the data becomes better at predicting what new shoppers will find interesting.
Where this reaches back into the analog world is how good old-fashioned synonyms play a role. What one person calls a dessert plate another person may call an appetizer tray, luncheon plate, dessert tray, or acrylic plate; the synonyms are identified by humans, then tested within the system. Knowing to tag a product with such terms directs consumers to that product more quickly, before they lose interest. Called “tuned searches,” the company found that in three months they could increase sales $30,000 by tuning two dozen terms on a handful of products.
Helping the consumer more quickly find the product they want to buy increases the conversion rate. But the SLI solution can also increase the revenue per visit (RPV) by suggesting similar or complementary products.
But what about the disconnect between actually holding or touching a product, examining it from all sides, and truly experiencing it before buying?
Perez emphasizes that SLI offers a complementary function to Crate and Barrel’s in-house e-commerce team. “Crate and Barrel is known for product development, selection, and exceptional merchandising,” Perez says. “What we do is help turn their online shoppers into buyers, and then sell more to buyers.” Beyond Crate and Barrel, the firm’s predictive product discovery solutions are employed across more than eight hundred websites with customers on five continents speaking in twenty languages.
Driving the outsourcing of this function is the fact that just as tastes change, so too do methods and means of shopping. The migration of online shopping to mobile smartphones is already complete. The trend emerging now is visual search. “Consumers are beginning to point their cell phone at an object, for example a chair, and search for other chairs like it,” Perez says.
The nature of all of this is a continuous tweaking, as Perez describes it, of the solution configuration for Crate and Barrel and all other clients.
Michelle is still needed on the store floor; a significant market segment still loves to shop in person. But if that same shopper turns to the web, SLI is there to help enhance the customer journey and show that shopper exactly what they want to see.