The perils that befell K-Mart and Sears might lead you to think offline retail's best days are gone, says Kerry Liu. "But if you dig into the data, offline retail isn't dying." Other verticals – especially in higher-end, omni-channel retailers, bricks and mortar are a still key. "Ninety per cent of the spend is still offline."
Digging into online/offline retail data is what Mr. Liu does. He's the co-founder of Rubikloud Inc., a Toronto startup that specializes in retail analytics that cross-references datasets from companies' online and offline operations – operations which all too often aren't talking to each other.
"If you look at the top retailers in North America, excluding Amazon and Apple, most of them started off as brick and mortar companies," says Liu. "They have a legacy store world, then an e-commerce world."
That means that their IT systems were usually designed at different times, and for different purposes, and aren't always sharing data. This can mean information from systems like the all-important customer loyalty programs that track who buys what; from inventory systems that track what's in stock when; from point-of-sale systems; and from the e-commerce platform, which is oftentimes a world unto itself.
By correlating this information with its own database of geographic and demographic information, Rubikloud is able to tell companies about who their customers are. "There's a lot of companies looking at retail, and they're taking a weird approach. They look at lifetime value of a client. They look at price sensitivity. But none of it makes sense if you don't understand your customer very well," he says. "A new grad who lives downtown has a very different basket than a family who lives in Whitby."
(Mr. Liu says that parsing data from location-tracking technologies like iBeacon will also form part of their offering; all of their data, he says, is anonymized and contains no identifying information about customers.)
From there, Rubikloud is able to use the data to advise their clients on how best to manage their ad spend: In a world where customers aren't distinguishing between a store's online store and their store at the mall, what channels should they use to reach customers, and how much should they spend on each?
Rubikloud started out aiming to provide meaningful analytics to online retailers – but quickly discovered that they were able to easily import data from brick-and-mortar operations as well, and could deliver insights by cross-referencing them with customers' online behaviour.
Mr. Liu says the 13-person company is always looking for new applications for its data; he says Rubikloud can help companies escape from the "vicious cycle" of going to consultants looking for business advice, by giving companies a predictive modeling tool that lets them work with their own data on an ongoing basis.
"The problem with consulting is that all that data gets lost with the consultant. For us, all that data stays in the system."