Beyond HumanBig PictureCatalystsConnected WorldExchangeMarketing MixNew MoneyNew SchoolPeople SciencePulse

When objective data meets a subjective industry

Peter Stojanovic

Would you be surprised if retail returns data shows that men who buy coloured socks are more likely to return items? It's part of the industry says Clear Returns CEO Vicky Brock.

One of the benefits of using data is that its objective by nature.

The measurements you collect on your consumers or services can be quantitatively analyzed efficiently (hopefully) providing you with clear results to help with strategic decisions.

What happens though if you work in an industry that is inherently subjective – where number crunching can’t quite answer the nuances presented within, say, human-orientated scenarios.

Vicky Brock has found herself in that position.

As CEO of Clear Returns, Brock and her team have to statistically anaylze retail returns data and to help them reduce funds lost through returns. This includes a serious amount data mining, cleaning and presenting.

Yet retail, and the return industry, is centred around human interactions with their purchases, and as a result, they will have have different shopping habits, reasons for returning items, and so on.

Can data science really measure different people’s buying habits in a useful way?

“I think it does stretch the statistical cliches about correlation is not causation…there are more things going on in this situation.”

One example of this is a shopping habit Clear Returns has noticed in the male market: colored socks and return rates. Apparently men who buy pink socks are more likely to return their items.

The ideas of the correlation and causation raises interesting questions but Brock doesn’t want their retail returns data to get lost in the abstract.

“What people need to do is look at the product. We see some items have a return rate of 90% and, when you actually see it, you think ‘why wasn’t it returned at 100%?’”

That question is critical to Clear Returns’ remit, or, at least the ‘“why” part is. It’s here that the company can use their retail returns data and detail a product’s chances of getting returned and pass that back onto the retailer.

“Why – it’s the biggest area of contention and if people write what was wrong with a product they return, it rarely gets quantivity put back into the system. So our retail returns data includes a lot of data mining of reviews and feedback to get granular on why things are coming back; we tend to mix the subjective with the harcore data stuff.”