Big data in use…retail

The retail industry is a big data pioneer that could be said to have been deploying big data before the term was even coined. It has long used loyalty cards and CRM systems to capture and analyse information on customers, surely an early take on big data?

But the advent of ‘big data’ takes that concept much further, delivering to retailers unparalled insight into their customers’ preferences, behaviours and purchasing habits.

In the latest of our ‘in use’ blog posts, we look at how retailers are using big data, and what is required for a retailer to truly benefit from big data. 

Retail big data – providing a personal touch
Pre-internet, it was said that good customer service in retail was based around knowing your customers. Whether that’s a tailor knowing their clients’ suit measurements or a grocer knowing what someone’s weekly shop entailed, understanding and anticipating customer needs and showing a personal touch when interacting with that customer were all of the highest importance.

Those basic principals still hold true today.

Starbucks and big data
Data can be deployed in any manner of ways too. For example, Starbucks has used data to help define its growth strategy for future store openings. In 2008 the company had been forced to close hundreds of its stores and when the time came to open new stores again, Starbucks took a data-driven approach.

It analysed enormous volumes of existing data relating to planned store openings, factoring in location-based data and demographics to determine the very best places to open new Starbucks stores, without damaging sales at existing Starbucks locations. So big data not only helped plan when up to 1,500 stores would be located, it also contributed to driving revenues for those stores.

Big data needs a bare metal cloud to flourish
This personalisation works even better online, where there is more data with which to work with. So in times of rainy weather, a department store could not only target customers with umbrellas, but offer umbrellas that match with outfits the customer had bought from that retailer previously.

But the key to extracting value and ROI from big data is in the ability to analyse such data in real-time - that is what gives retailers the opportunity to make such targeted and tailored offers and services. But big data always demands performance and the virtual enviroment favoured by many retailers is simply not powerful enough.

Deploying Hadoop or other big data applications brings powerful analytics to a retailer, but doing so in a virtual environment will see performance suffer and may prevent the real-time analytics that is so full of potential. Many industry experts have stated how hypervisors are a major drain on performance. This is why we have removed the hypervisor from our big data infrastructure, allowing a retailer to benefit from the full performance and power of bare metal.

Some industries are better suited to certain technologies than others. Retail and big data is one such example. It is an industry that can arguably derive more value from big data than any other. But the key to doing so lies in a retailer’s ability to move and analyse this detail in real time and the key to this is computing power.