Online Retail’s Big Data Opportunity

Using the right consumer research to close sales

The era of Big Data has arrived. But in a recent survey reported by eMarketer, more than half of retail executives were either not clear on what Big Data is, or not sure of its implications for retail.

In retail, Big Data is created from many consumer touch points both in store and online. From hundreds of thousands of online clicks, transactions, or ad impressions you can gain insights into what consumers like, the exact type of product they order most, the prices they pay, where they come from and where they leave to. We’re not talking new data per se.

But because it is the interplay of all these signals that creates insights and value, it doesn’t readily fit into the field structures and taxonomies of traditional database models. And it is indeed big, easily exceeding the handling capacity of traditional data tool sets.

Big Data also has velocity—not just in the speed of its creation, but the speed of its obsolescence as well. Unlike CRM, it’s not something to be crunched offline and subsequently applied to e-mail and catalog targeting. Harnessing its value requires the ability to act fast.

What is it all good for? Decisions. In particular, real-time decisions on what to show a given visitor to improve the relevance of marketing and site experience.

For example, to whom do I need to show a promotional offer in order to drive a conversion? Or, perhaps more interestingly: who does not require a promotion to convert? From that initial binary decision, it’s possible to gradate your promotions so that people are only exposed to the level of promotion necessary to motivate a purchase.

Another example: de-averaging my site experience to improve the user experience (and capture more value in the process). Typically, retailers create a “standard” site experience, with emphasis on moving visitors into the cart. But as we know from a decade of 3 percent average conversion rates there are going to be lots of cases where the retailer should offer a different experience. Perhaps one that emphasizes the offline stores or seeks to monetize the visit with advertising in the absence of a potential product sale.

Should I push my loyalty program to visitor X? E-mail registration to visitor Y? Latest product releases? My closeouts? New store openings? My St. Jude fundraiser? How-to content? In the competition for scarce site real estate, a big data predictive model does a better job than gut instinct or political muscle in deciding what gets put in front of a given visitor. Because the answer is different and varies by visitor, it’s no longer an all or nothing decision. Visitors see the options most relevant for them.

Retailers like Amazon and Walmart have the wherewithal to push forward aggressively with internal resources. But for everyone else, third party providers offer the data science skills and predictive analytic capabilities necessary. And despite all the hype, Big Data isn’t the latest vendor conspiracy to get retailers to re-build their entire tech stack. Decision-making capabilities often can tie in directly to existing systems so that your current CMS, promo-engine, ad server, etc. can do the work.

For retailers who get a handle on Big Data the payoff will be significant. McKinsey estimates that retailers could boost operating margins by 60% by effectively harnessing Big Data, with the bulk of that coming from marketing and merchandising.

As former LinkedIn data scientist D.J. Patil told Inc. Magazine, “Data can become the weapon of choice by which you make your stuff really work.” For retailers, it’s a weapon that is just beginning to be deployed but will have a dramatic impact on future performance.

Rob Schmults is Senior Vice President of Strategic Partnerships for Intent Media Inc.

Related Content