If you spend any time reading the technology media, you’ll probably be familiar with the term ‘data is the new oil’. It’s a way that tech experts have chosen to frame the value locked up in the information that increasingly data-driven businesses gather from their activities and customers every day. It’s chock-full of insights but how on earth do you separate the useful bits from the background noise?
This is a conundrum we’ve spent a lot of time puzzling over at Dressipi. There is undeniably a perceived value in having lots and lots of data – for example, we now have the fashion fingerprints of over two million women. But none of this data means anything if we can’t use it to create services that benefit two quite distinct audiences. The women who use our style-recommendation services want advice that’s tailored to the data they provide us. And the retailers who pay our bills want us to create actionable insights on how Dressipi’s users shop for and buy clothes online.
Being a business that serves two audiences – but that makes a living from just one – raises a number of questions. The first is ethical. Our end users entrust us with some very personal information. They tell us about their height, their weight, their body confidence and, in order for them to trust us enough to use our services, we have to keep that data safe. So that means we never share any information about individual users with anyone – not even the retailers who pay us for our services.
Moreover, even if individual Dressipi users did give us the permission to hand over their vital statistics to retailers, they wouldn’t learn much. In our experience, looking at how individual customers behave can be valuable in providing more and more personalised services. Whilst looking at individual customers can be interesting when it comes to creating actionable insights, there is always the risk that person is an anomaly. So it’s actually much more important to examine overall trend data.
For example, one of our retail partners initially chose to work with us because they were finding many women were buying two items and then returning one of them. Naturally they assumed the problem was because these customers didn’t know the right size. When we crunched the trend data, however, we learned something different. While 15% of the garments returned to this retailer were due to sizing, 55% of garments were being returned because customers were buying two similar styles and then sending one back. This told us that, for this retailer, the key to reducing returns was not helping women select the right size but providing better style advice up front.
In this new information-rich world, data may indeed be the new oil but in our experience the real power lies not so much in having the raw materials as the ability to refine it. We’ve used data to challenge long-held assumptions about the fashion retail business and help our customers overcome what they thought were intractable problems. Along the way, we’ve also helped hundreds of thousands of women to shop more confidently and efficiently. Yet we’ve only been able to do this because we view the collection of data as being just the beginning of the hard work. 
