IBM tunes big data appliance to retailers

IBM (s ibm) is working the reins of its Smarter Commerce initiative by rolling out a new Netezza analytics appliance designed to help retailers churn through petabytes of consumer sales data in real time. With the IBM Netezza Customer Intelligence Appliance, the systems giant is trying to capitalize on its contention that online activity is becoming increasingly important for retailers, so they can’t afford to not understand customer behavior and improve the web experience.

According to the IBM press release, “70 percent of a customer’s first interaction with a product or service takes place online,” which underscores the need for analyzing customer activity to find out what they really want. If consumers connect with a retailer’s site and enjoy the experience of shopping there, they’re less likely to spend their money on any of the countless other sites selling the same goods.

Of course, many large retailers — especially those that can afford IBM Netezza systems — also have brick-and-mortar stores, and IBM isn’t omitting those operations. The appliance actually aggregates data from in-store and mobile shopping, as well as online shopping, to let retailers connect the dots between activities across the different channels. In theory, all this information should help stores create better promotions, target specific user types, and perhaps even optimize website or physical-store design.

The new Netezza appliance also incorporates business-intelligence software from IBM’s Cognos line and specialized retail analytics software from IBM business partner Aginity.

When IBM talks about the new appliance handling petabytes of data, however, it’s not talking (at least directly) about the type of unstructured data that has made Hadoop such a hot technology in the big data space. Essentially an enterprise data warehouse, Netezza is SQL-based and therefore ideal for structured data such as customer records, whereas Hadoop is all about unstructured or semi-structured data coming from sensors, social-media streams, server logs and other non-traditional sources of business data.

One step on the path to big data nirvana is combining structured and unstructured data to gain an even more-complete view of what customers are doing and feeling, which is something IBM is trying to do through tighter integration of Netezza with IBM’s Hadoop-based Infosphere BigInsights software. It’s not alone in this quest, as rivals Teradata (s tdc), EMC Greenplum (s emc), Oracle (s orcl) and HP (s hpq) all have similar strategies in place with their respective components.

However, if revenue growth has anything to do with how IBM and other vendors continue to expand their big data capabilities — and it does — we’re only seeing the tip of the iceberg in terms of what will be possible. In IBM’s fiscal fourth quarter, business analytics revenue was up 16 percent year-over-year, Smarter Commerce-related initiatives were up 25 percent, and Netezza revenue jumped up 70 percent.