I’m a big fan of Big Data in the Retail space. Retailers of all sizes have already chosen Sisense for their complex data problems, from Mejier to Target.

These analytics leaders understand what’s required in order to manage complex data at scale. Any retailer sees the opportunity to market online and at their location but do they have the data systems that can grow as fast as their business does, without requiring more investment in IT or in human capital?

As Mike McNamara highlights in the below video, technology needs to enable you to differentiate yourself, regarding your size. Mike knows what he’s talking about – Tesco has over 16M loyalty card customers, is the world’s 4th largest retailer in the world and can probably predict, before anyone else, when a recession is about to it. Here are some best practices I took away from my interview with Mike.

Timeliness and Accuracy

When implementing data systems, remember that the the type of report you serve matters. While flashcards need to be provided quikly, they only have to be directionaly accurate. Operational reports, on the other hand, have to pivot around accuracy. If your analysts have to generate 100,000 variants of coupons – the data they use better be accurate!


Like in any business, the devil’s in the details. Storing aggregate information is not enough. Your data infrastructure needs to accommodate millions of rows of data, sometimes terabytes of data. Additional, this data needs to be accessible by any of your people (not just data scientists). Todate, most companies have had to build two technology layers: one or several data warehouses for data storage and one or several business intelligence tools for agile analytics data exploration. Datawarehouses can be built using Microsoft, Oracle, MySQL and others. Agile Business Intelligence vendors come in masses, from legacy in-memory vendors to newer visualization tools.

Well, welcome to Sisense – the only Big Data Analytics Company that provides all you need in one package – from a columnar datastore build for scale, a disk-based in-memory engine built for speed…and visualization built for normal people. Find out more at retail analytics.

Tags: |