SKUtrak: Managing the complexity of different customers

Supermarket trading data represents a tremendously valuable picture of shopper demand.

However, creating a clear image from the myriad component parts presents a hugely complex problem. Grocery retailer datasets that seem similar in principle turn out to be very different in practice; different formats, structures, terminology, measures, definitions, product codes and descriptions, collection methods, granularity to name just a few.

They differ so much that once you explore them in detail it's hard to establish where the similarities lie.

Many organisations have turned to "Big Data" technology to address this inherent complexity. Large supermarket suppliers around the world have implemented Hadoop platforms and data lakes to address the challenge. These approaches certainly offer solutions to storing wildly varying data but provide very little in terms of reducing or managing complexity; either end-users need to know the source data systems in mind-boggling detail, or your technology team arbitrarily reduces complexity by combining seemingly similar data.

The results vary from irritating to catastrophic; from thousands of hours of wasted effort, as different teams interpret data in different ways in different tools, to badly combined data powering critically flawed business decisions at huge opportunity cost.

There has to be a better way... and there is.

SKUtrak combines deep grocery supply chain knowledge with an essential mix of critical technology components that together reduce complexity with certainty and provide a rich, reliable, curated data source for your decision-makers.

At its core, SKUtrak contains a robust data model which has been validated against the operating practices and trading data of multiple world-class grocery retailers. This data model represents the Flow-of-Goods through grocery retail in a flexible, practical, robust and consistent manner. SKUtrak translates retailer system data from its native (retailer-specific) form into this model in a reversible manner; as a result, you can treat customer data as if it came from a single system and still present information in customer-specific terms where required.

In addition, SKUtrak provides master data management tools that automatically suggest matches between your product definitions (codes, descriptions, units of measure etc.) and those in your customer's trading data. Your data owners and administrators simply approve or amend match suggestions and then rely on SKUtrak to provide trading performance information across customers, using your product perspective, or customer-by-customer using theirs.