”… and suddenly, as if by magic, the shopkeeper appeared!”
Talking to an ex-colleague recently, the subject of behavioural analysis came up, provoked by an intriguing little habit Facebook’s got into recently, where it’s been advertising merchandise to me from some of the bands in my music collection. The crux of the issue had two parts. Firstly, and quite simply, what does one do with all that data? Obtaining it and storing it is enough of a challenge, but how do you go about making sense of it and using it to drive a business? Secondly, isn’t this all a bit of a black art? Don’t you need highly-skilled practitioners to make it all work?
He could clearly see the benefits from a business point of view (increased conversion, reduced cost per acquisition, improved customer experience etc), but where the difficulty lies for him is in the operational impact this would have on his (small) team. In my experience, it’s not an uncommon concern – many medium-sized retailers do not have the latitude to hire in new teams to support a data-heavy application, and so view the operational side as a massive challenge. However, perhaps the concept of behavioural analysis isn’t quite as new-fangled as the technology would have you believe.
In looking at how behavioural analysis works in principal, I’m reminded of my experience of running an Oddbins store in a medium-sized Cheshire village, quite some years ago. One of the unwritten rules of being a store manager or assistant manager at the time was that you would be expected to be able to remember at least 100 customers’ names and buying habits, and most of us could probably do more than that. So, when a customer wandered into the store, you’d recognize them, greet them, remember what their buying patterns were, and adjust your approach accordingly – even to the point where you knew who liked to be left alone and who usually wanted a chat. (I’d also like to take this opportunity to apologise to one Mrs. Talbot, who repeatedly asked me to stop selling her husband so much Burgundy, and whose entreats I steadfastly ignored. Sorry Mrs Talbot.)
This basic sales technique (less a ‘technique’, more common retail sense) could also be extended to customers you didn’t know. If someone’s scrutinizing the top shelves of the Bordeaux section, having parked the Bentley outside, you’re not going to start waxing lyrical about the new £5 Aussie Shiraz. If someone in formal dress rushed through the door in a panic, you’d be thinking ‘emergency gift’, not ‘wants to be talked through the range in excruciating detail’. This is behavioural analysis at its most basic, and like it or not, whenever you walk into a store, a good sales assistant will be going through this process.
Extrapolate this for a moment to the majority of online retailers. Whilst you’re browsing their site, they don’t know what you look like. They don’t know what you’re looking for. They don’t know what mood you’re in, and they can’t make any inferences about any of those things. They will sell to you in exactly the same way as they’ll sell to every other one of the thousands of other people browsing the site at that moment. To me, this sounds like everybody loses: the customer has a sub-optimal experience, and the retailer’s normally well-oiled sales machine is bound and gagged.
So what if certain bits of information were available to the retailer at the point at which you enter the site? Perhaps if that site knew that, as you’d clicked ‘Fine Wine’ first in each of your past four visits, now might be the time to advertise the arrival of a new and particularly good value parcel on the homepage? Or have you come from Google, where you typed in ‘wine gift’? Should that site be advertising gift sets from the off? If you’ve been researching that fridge for several weeks, is now the time when a small incentive might accelerate that purchase decision?
Granted – a blunt implementation of this can be annoying, as with the Facebook example. I’m the wrong side of 30, and stopped wearing band t-shirts about fifteen years ago – this is a good example of poor, one-dimensional targeting.
So, it’s fair to say that the fundamental principles of behavioural analysis are ones that will be instinctive to any seasoned retailer, and that, for most businesses, the data necessary to drive such a strategy exists. The missing links are simple: a sensible way to expose that data to a relevant application, and an interface to that application that makes sense to a business user whose core skill is selling, not data management.
The beauty of it is in the marriage of old-school retail sales skill and the sensible application of business intelligence technology by people who understand retail. We’re currently performing this service for a number of our retail clients – if you’re keen to see how behavioural analysis can transform your business, drop me a line at dan.wilkinson@conchango.com.