Predictive personalisation can be interpreted by everyone to mean something different. It can be daunting sifting through 400+ email service providers (ESP), all chattering on about personalisation, let alone be able to make sense of what the distinction between each is. It was about ten years ago that the lights began to go on, as people realised each consumer bought more if you what we offered was pertinent to them.
Back then this appreciation was believed to apply to use of just a few elements, like the consumers [name]. Indeed it is a sad indictment today that there are still so many who haven’t moved on from this. If you Google email personalisation now, and you’ll find even the search engines struggling to offer clarity among, what has become the real principals of personalisation. So many people still regurgitate the spiel heard at university, banging on about it being name, age, geography, gender etc.
So what about product preferences? Hang on, that’s just opened another can or worms. As soon as we discuss product content the next word you hear is segmentation. But a segment is not personal. If you, as one person on an ecommerce database, bought a pair of jeans in the last 3 months, the adopted inference is that you have a positive inclination (and the money / propensity) to buy another pair now.
It doesn’t matter what else you bought, tops, coats, boots, socks, jumpers indeed anything; you just get lumped into a corral of people that bought that item too. There is little if any correlation between the two, just sliced and diced database. This is just about the furthest you could get from analytics, but because time is money, there are deadlines to meet, and a CTR target to achieve, off you go. A missed opportunity of utilising the nuances and subtleties of the data your platform gathers every day, lost.
So where and how does predictive personalisation come in?
The main reason retailers continue to use segmentation is because it’s easy. It is not viable for even the most competent marketing staff to have the time to drill down to each individual customer, especially if you have millions of people people to look after. It can’t be done, or so you may think, until predictive personalisation came along.
Segmentation has now been replaced by a 1:1 level personalisation. However, the boffins quickly appreciated that it wasn’t just the products themselves that had a measurable sequence. The metrics track both products purchases, and the sequence of navigation used to arrive at that purchase. The amount of time each SKU is viewed, frequency hits, and individuality of that whole sequence, make up the structure of the buying decision itself, and has become the focus.
This has fragmented each discipline into specialist pockets. It has become lucrative to be a data-analyst. Data management is now big business. Predictive personalisation, through the advance of technology, and by its own success has, just like cells, become sub-divided.
Before we explore this, we should just make a distinction here. Another popular topic when dissecting the advent of predictive personalisation, is content. The platforms themselves want their pound of flesh too, as it offers them the biggest voice, and additional revenue.
But by doing so, presumes the consumer is locked-in. However loyalty is not guaranteed. None of us go around with VR goggles on, only looking at your fav website. There is a small matter of going to work, sleeping, family, friends, health, car, mortgage, kids, TV, holidays, sport and of course food to fit in. So the only way to consider predictive personalisation in its purest form is where it actually meets the consumer.
It must involve going to them, to capture, entreat and nurture their attention away from all the myriad of interests clawing them away. To invite them back to your site to buy, again and again. The degree of your success of this has, and remains, the most effective focus of all ecommerce marketing disciplines. It is the leading provocateur of interaction, and therefore one that enjoys familiarity, which consumers verify by subscribing to your site.
How to make predictive personalisation work for you
Predictive personalisation software (PPS), is typically a plug-in that once installed on your platform acts totally autonomously – without any human involvement whatsoever. If you involve humans they have a habit of detracting from the purity of the data, which is telling you exactly the perfect products unique to each individual, and the time to offer it, again unique to each one.
Gone is sending out 50 segmented emails a day. With PPS installed on your site, the AI/ML algorithm is already delivering the maximum return possible. It doesn’t need to be adulterated by adding left field interest products too. When you start doing that the purity and clarity of each selections made for the individual go unnoticed, and denigrate the purpose. More importantly it lowers the return.
Consider your own whims for a moment. If you have looked at a product a couple of times, to consider something you’re thinking of buying, and then as if by magic, that very item lands in your inbox on pay day (or as the algorithm sees it – the day you are most likely to make a purchase), you’d be impressed. That’s exactly what happens with everyone else too. The effect is so prolific, the stats suggest the return on investment is 20x higher than email marketing and omni-channel marketing combined.
When installed it negates the need for any necessary staff cost. Installing PPS allows you to have a catch-all, so you can move on to getting new customers instead. More customers, then feed into the product process, with the highest return, being fed autonomously by PPS capturing the maximum amount of sales – the perfect cycle.
Predictive product selection software maximises the ROI, increases basket size, and delivers the highest rates of customer lifetime value. Therefore your first and only marketing tool is PPS software.
SwiftERM are a Microsoft Partner company. Enjoy a month’s free trial of PPS software here. Keep all the profits – no lock in.