Don’t listen to the boo-birds saying “big data” doesn’t work. It does — if you have the right tools.
“The quickness to market is incredible,” Gregory Phillips, a senior web analyst at OfficeMax told me yesterday. “It’s literally hitting a button.”
I spent some time chatting with Phillips and a few executives from marketing optimization firm Monetate yesterday about a new product that Monetate is releasing in public for the first time today: LivePredict. It’s an automated web optimization service that identifies sales opportunities for an almost infinite number of dynamic customer segments and targets them instantly with customized offers.
And it’s boosting sales in beta tests by almost 30 percent.
Big data, meet big sales.
The software-as-a-service package, which comes as part of Monetate’s full marketing and site suite, works continuously to identify large or small slices of the market that are responding well to a campaign, turning standard A/B testing into massively multivariate testing automagically. As a result, marketers discover demographics they never knew existed — and target them with tailored offers in real time.
In one campaign, OfficeMax discovered a huge response to a particular set of products from people 19 to 24 in just one region of the country. It was something “I never would have thought to go dig for,” Phillips says, but OfficeMax was able to immediately take advantage of it. In another campaign, an outdoor apparel company discovered that “married couples who work” was their single highest-performing segment, Monetate VP of engineering Tom Janofsky told me. Again, it wasn’t something they knew already despite having a few hints, but it enabled them to immediately begin his-and-hers campaigns.
LivePredict takes four broad sets of data and correlates them in real time: technographic data, demographic, data, geographic data, and a company’s own customer data, behavioral data.
The technographic data — devices, mobile versus desktop — gives marketers info on average orders on iPads versus Android smartphones, and so on. The demographic and geographic data supplies information such as region — fairly basic — but also up-to-date info such as weather, so retailers can target beach towels to the heat wave states and snowblowers to Minnesota, and not vice-versa. In addition, LivePredict integrates the typical demographic data points: age, education, income, and so on.
Add the company’s own customer, sales, and order data in real time, plus load-up marketing campaigns, and LivePredict is ready to start telling you where to sell what for the greatest return.
“Amazon changed the whole online experience for the consumer, nailing the personalization thing,” Monetate exec Marifran Manzo-Ritchie says. “That was the promise of e-commerce: the right product in front of the right customer at the right time. People don’t want to be looking for women’s shoes and see children’s clothing … merchandisers should be adapting their site to what people want and need.”
That’s the kind of customization that could never happen without technology, of course. In the past, retailers would run simple tests with simple variables, according to Janofsky. A third would be heroes, a third zeros, and a third inconclusive.
But not any more:
“It’s really the logical next step for the combination of segmentation and testing,” Janofsky says.
VB's research team is studying web-personalization... Chime in here, and we’ll share the results.