Data is a very vague term, particularly in the world of retail operations.
With retailers now collecting such a vast amount of data points, it can be difficult to translate them into trends and funnel these findings to strategies. The key is to first look at data sets separately and then find how to connect them in complementary ways to best serve consumers. For startups and digitally native companies, this is the foundation from which they operate — the ability to collect information and use it to evolve rapidly is what keeps them standing.
At SXSW, Stitch Fix chief algorithm officer Eric Colson and Poshmark co-founder Tracy Sun discussed how their companies are specifically using two types of data: explicit and implicit. As two e-commerce companies that exclusively operate online — Stitch Fix is a subscription clothing and styling service, and Poshmark is a marketplace for resale fashion — they are using both types of data to better connect with consumers. Here’s how that looks in practice.
What exactly is explicit data?
Explicit data is any information a consumer actively provides on their own. Common examples include their name, gender, email address and home address, and they’re typically offered up during a customer’s first transaction, or when signing up to receive a newsletter or other information through a site. From this data alone, companies are able to tailor settings to provide options that would likely be of interest to particular shoppers, such as spotlighting popular styles they can pick up in their regional area.
“Explicit data is the preference data that consumers tell us about themselves, about style and fit,” Colson said during a panel at SXSW. “This is very specific info provided to us by the client, such as ‘Here’s my size, and I don’t love skinny jeans.’ This creates a copacetic relationship between us and the consumer. They’re trying to help us help them.”
Got it. So then, what is implicit data?
This is information retailers glean from a consumer’s browsing activity on their sites, like what styles and brands they’re clicking on most and what parts of the page they tend to hover on. “You don’t necessarily realize that you clicked on that Gucci cross-body bag, but we know because you clicked it two or three times in a row,” Sun said. “We know you’re pretty interested in the Gucci cross-body, and we should show you that again, and show you things like it or the [brands] who are showing things like it.”
And how specifically do retailers use this data to target shoppers?
By using the two together, retailers can help curate styles shoppers may like and direct them to the brands they’re most interested in. It also allows sites to make recommendations, Sun said, and direct shoppers to areas of the site that will be most helpful to their individual needs.
This has allowed Stitch Fix to implement programs like Frankenstyles, garments that are created using data algorithms — the process calls to mind Darwinian evolution: Using a survival of the fittest model, the most popular styles merge together to form new styles. For example, due to their popularity, a ruffle on a skirt and a pattern on a blouse may later come together in an altogether new garment.
How much data is the average online retailer collecting?
That depends largely on the retailer. For a company like Poshmark, which incorporates a social media element by allowing sellers to connect with potential buyers in an online community setting, this can be up to 1,000 data points each day, said Sun. “We know what you’re looking at, what you’re liking, what brands you like. If you’re looking at Gucci for example, we know that you’ll also probably look at Louis Vuitton items. For us, it’s a people connection, as well as a connection to inventory.”