Why retail is failing with big data?
Posted on Thursday, August 24, 2017
When it comes to big data, retailers are often presented as being at the forefront. Programs like the Tesco Clubcard loyalty scheme have helped the British supermarket better understand its clients – and to boost sales.
But the reality is that retailers (and other big data users) are not yet able to take advantage of their datasets or analytics in real time. Most simply do not have the infrastructure required to collect, analyze and report on data gathered at the checkout in real time.
The Cloud is not the solution
Cloud systems are often presented as the ultimate answer to almost every IT challenge. But when it comes to real time use of data, the cloud is always found lacking. The latency inherent in internet traffic makes a significant difference to access speeds for instance.
There is also the cost factor. With data constantly flowing in and out of a cloud service, it becomes almost impossible to properly manage resource usage – and therefore spend. Retailers may end up spending more for lower performing systems if they assume the cloud holds the answers to all their big data problems.
The Amazon factor
Analysts have noted that retailers are becoming much more circumspect in their choice of cloud service. Having seen the effect of Amazon on bricks-and-mortar outlets, retailers rightly regard Amazon as a significant competitor who should not be holding on to their sensitive business data.
This means they must adopt Microsoft Azure, or take a huge risk with Google services. Either way, the choices are limited.
Playing it safe, keeping it in-house
The performance advantage offered by on-site systems is significant – and cloud platforms show no signs of catching up any time soon. This means that retailers will have to build on-site storage infrastructure capable of delivering real-time analytics to the edge of the network, or lose ground to their competitors who do.
Big data will still be a big deal, but the cloud may not be as important as initially believed. At least not in terms of real-time analytics.
To learn more about how to build big data-ready infrastructure in-house using your post warranty storage arrays, please get in touch.