The past 5+ years has seen a proliferation of targeting and tracking capabilities, trying to predict customer behaviour and target them with the right content, offers and products.
But this isn’t a clear-cut science and as the more predictive our world becomes the smaller it also becomes.
Can you imagine your merchandising executive suggesting you cross-sell patio heaters with men’s bikes on a hunch? What if they said they had proof customers shopped both together? How miffed would you be to see them cross-selling a single mattress against a double bed? How about if a trusted intelligent merchandising application suggested these to customers, what’s your reaction to this?
Fix the application? The data? What if the tool had been doing its job. The correlation between bikes and heaters was there. Customers were actually buying these together.
Sure, we could tweak things to ensure cross-sells are more relevant, but surely this is how customers shop so we should embrace this?
Maybe. Or maybe not…
Taking the second example of the mattress and divan will help explain things better. If customers start to buy two incompatible products together, and other customers then also buy the same items together, you create a perpetual filter bubble. The more customers that feed into the data the stronger the correlation. The system thinks its onto a winner, the numbers prove this is working, sales increase. That is, until the complaints also increase.
The point is, if we always think customers know what they want, or we can predict their future actions from their past, we’re mistaken. Sure this is not new ground and we only need to think back to two leaders that changed their respective industries by not listening to what customer demand: Henry Ford and Steve Jobs. Both had a vision of what they wanted to bring to the world which would change their customer’s experience in a profound way.
There is a vast expanse between product recommendations and business development but both ultimately pivot around an end user. Whether it’s a service or product the threat for Personalisation is in losing that “wow!” factor – “wow look at this machine that I can drive which will replace my horse” or “wow this portable digital music device is amazing, I’ll ditch the tape” – and in a similar way we’re losing the ability to serve a similar experience to our bricks and mortar stores. If the next website you visit is 100% personalised to you, where’s the inspiration? If the last 3 cars I bought were Ford Fiestas a simple predictability model wouldn’t be incorrect in assuming my next purchase is either going to be a Ford or more specifically a Fiesta. But what if I’d just had my 4th child and needed something bigger or had just received a huge payrise and wanted something flashy? How would your website know that? It wouldn’t. At least not right now…
Taking circumstance into account when predicting and surfacing the right content, offers and products for customers is and will remain the most difficult part of online commerce. The human experience sets apart the best retailers. I spent 5 times as much as I originally planned a few years ago in London’s most popular Toy department store, Hamleys, due to their exceptional shopfloor team. They were passionate, knowledgeable and took the time to understand my needs – not recommending items that other customers bought, but specifically what they believed would be right for me (my kids, nieces & nephews to be precise!). And they were just that, perfect.
And so the digital world is still evolving. We’re yet to connect all the dots. There was a discussion years back around Web 3.0 and how the web would know you. From the moment you searched, to the moment you landed on a retailer’s website. It would know your preferences, your marital status, income, work status and more to offer a completely unique, tailored experience. Forward on to 2013 and think where Facebook is taking us… as a marketer I’m all too aware the difference this level of data could provide yet as a consumer, I’m not yet willing to give this data up… so where does the future of personalisation lay? How can we maintain that wow factor?