Big Interview: Head of analytics at JD Williams on using data in retail

'Big data' is often described as just another technology buzzword. But JD Williams is most certainly big on data. The catalogue and online retailer creates 65GB of customer data from its website every month, which is used to gain insight and deliver business change.

Gareth Powell, head of web analytics at JD Williams, explains the retailer is using its data to optimise its marketing budget using predictive analytics and forecasting.

He says since implementing software from Celebrus in 2010, JD Williams has been able to capture data from each time a customer visits its website. "Rather than a holistic view we wanted a granular view of the customer," explains Powell.

The retailer is able to stitch together customer insights by using Celebrus technology as well as Teradata for data warehousing.

Celebrus is deployed using a piece of Javascript which allows JD Williams to capture information whenever a customer navigates the website. "The average session is 12 minutes long, so we're capturing a heck of a lot of information as the customer moves through the site," he adds.

Marketing optimisation

The retailer has been using this information to make changes within the business. One such change is understanding customer tastes in order to decide which customers receive a full sized catalogue in the post and which receive the reduced size. This optimises catalogue production and spending, leading to a 3.7% improvement in overall revenue.

Another marketing optimisation decision based on big data is to eradicate inefficiencies in paid search spend by analysing the business impact of different keywords. Powell says understanding the value of TV advertising is also important to decide whether investing in peak or day-time TV brings in more customers.

"There was also one surprising insight before Christmas," describes Powell. "Because we understand what devices customers are using to visit the website, we discovered if a customer is using three devices, they're worth 300% more to us than someone using one."

Powell explains that his team realised there was a population of customers who accessed the website by desktop, but not tablets. To capitalise on this, the retailer offered a discounted – almost cost price – tablet to customers for a mere £35.

"It was a real resounding success," he said. "And we sent out between three and four thousand."

"It gets them all using that technology and if we can get more people on that platform it leads to more sales and more happy customers."

Bricks and clicks

Because JD Williams is a catalogue retailer with a digital-first strategy, it is lucky enough to track its customers fairly easy using their account number. "That's a good start from a customer data point of view because high street retailers don't have that visibility. Some gain it through loyalty cards and can then stitch that online and offline data together."

Powell continues to say that pureplay retailers with no legacy systems and stores do not have the same challenge of stitching together the data.

But as part of parent company N Brown Group's wider strategy, it plans to ramp up its physical store presence eventually reaching around 80% of the population. But with stores comes difficulties with data analytics.

"Then we've got a similar challenge to the high street where we can't stitch the data together," he adds. "The store sales are a small proportion of business sales, but they have a halo affect, which is measurable but difficult to allocate to individual customers."

JD Williams is one of N Brown Group's three core brands along with Jacamo and Simple Be, which both target younger customers. Powell says while the structure internally has three teams, learnings from analytics and digital are shared. "We're all working in the same space, and the analytics team supports the three areas – we may have a specific learning from Simply Be we can take to JD Williams."

Powell says along with tying up store data, the next project for the web analytics team is website personalisation, where he wants to be able to present a unique experience to each customer.

"It will take some time, but we're on that journey."

The data journey

But that journey is not easy and when the retailer started taking data analytics seriously Powell says there was a pain period. "It's like trying to see the wood through the trees, there's so much data – what do we do first?"

"But we've been on that journey since 2010 and we've understood how to exploit it and ultimately how to make actionable changes to the business," he comments. "And it's a really important piece to our business now."

He says he understands how other retailers become so overwhelmed when it comes to embarking on a data project. But Powell warns that those retailers who are not analysing data are going to be left behind. "If they're not capturing, configuring, digesting the data, that really spells trouble."

"When I think of the retail failures in recent years – Jessops, Comet, HMV – partly it was due to their inability to go fully digital, but also due to a lack of focus on the customer data."

"Everyone is fighting for the same pound – especially in fashion retail – and if they're not harnessing the data and understanding what you can do, it spells disaster."

Click below for more information: