Consider all the time, intelligence and effort that has gone into devising an AI driven email personalisation solution, that delivers the greatest possible return for your marketing investment. Would you then, randomly build in means to negate all that hardwork and potential? Why would you? And yet, there are solutions out there devised to beguile the unwary ecommerce retail owner, that obliterates the true potential ROI. This article considers those culprits and interlopers.
Segmentation
Segmentation by it’s very definition lumps groups of people together, as it serves are means to address a defined audience to a particular message, when you assume all the people in that segment fit an appropriate level of likely response. You don’t know for sure they aren’t likely to buy the particular product you want to promote, and perhaps stats from previous emailings have illustrated that the level would be sufficient to make their inclusion viable.
Consider then if someone one who didn’t know you, as well as they should, called you “you people”. You stop and look around, and realise you have been lumped in with a bunch of people with whom frankly you have very little in common. And yet within a marketing department, this argument is arbitrarily decided everyday, as they haven’t the time, inclination nor budget to care that much.
After all, if you’re offering a discount for example for the early purchase of a new line, you need it to go to a sufficient number to make it worth your while. Great for email service provider software (ESP) for promotional purposes, but what of the effect of the treatment of individuals as unique? Unscrupulous wordsmiths use words like ‘precision targeting’, deployed to infer the pinnacle of sophistication, while the reality is actually herding so it fits a purpose.
What if instead, you could offer the individual predictive personalised content, perfectly identified for each individual consumer by their own actions (buying history, impressions, navigation, length of stay on product, number of returns to that item, reaction to previous offerings etc). Instead of corralling them, what if your actions defined an individual selection exclusively for that person. What do the stats say about the potential for that solution? No wonder research houses, such as McKinsey and Forrester, have identified that the ROI is 20x higher.
This then is proper predictive personalisation, not obfuscating jargon, regurgitated to make old-hat solutions appear as predictive email personalisation.
Build-your-own variants.
If you have sophisticated AI/ML software, what on earth would provoke you to move away from precision product selection, to a world that offers human-beings the opportunity to cock it up? Yet this is exactly what some supposed predictive email personalisation software does.
If a human being is involved you know without question, that the first thing that has happened is to incorporate all those habits and frailties human beings bring with them, not just errors and omissions, but catastrophic ones at that. Further, before your campaign has even begun, you have invited all the inherent costs associated with those staff, necessitating a return before it even goes out, having already established that the return will be lower than predictive email personalisation software achieves.
Hopefully the logic of avoiding this insanity is beginning to be appreciated. True predictive personalisation email software does not need marketing staff to interfere with a calculated product selection for each individual in each of their own emails, it is already perfected to maximise return.
The common argument being, of course, is that a pre-format stylesheet takes away the ability to massage the design, and influence the look to seduce your recipient. What they don’t tell you is, that to allow that level of interference (designing it by hand) is directly proportionate to the reduction in your ROI. Literally twenty times what it could have achieved, making it a pyrrhic exercise.
Why are the sales returns higher? Quite simply precisely because it hasn’t been tampered with. These are unadulterated product selections. Your consumers aren’t stupid, we’ve already had enough experience of sales-hype first-hand. Instead PPS delivers their selection accurately selected for each consumer you have. It illustrates your care for that person as an individual, and in return for this personalized attention, they lavish you with their continued and increased custom.
Market leading recommendations
Another spurious reason to employ alternate poorly performing software is the supposed offer of “market leading recommendations”. Yaboosucks – so what? Consider yourself as the shopper. You like browsing, and are used to having ergonomic, easy to use, navigation systems to look and find the exact products you like.
These might include a comprehensive search facility, and clearly defined product definitions etc. What each of us are used to using, daily. ‘Market leading recommendations’ are just that, someone else’s. By their very nature the stats have already defined exactly what your customer is most likely to buy, who cares what anyone else likes, if it’s not their choice. What makes someone else’s selection more important than their own? This is particularly relevant to customer churn, or the reverse – loyalty, which grows. As the rate of return (RoR) of products almost dissipates with predictive email personalisation, which it doesn’t for imposter software.
Triggered solutions.
Just like ESP software, triggered solutions do have their place in the spectrum of solutions, sufficiently important to be included in your marketing stack. But triggered solutions are, by their very nature, a quid pro quo email personalisation solution, i.e. it requires something to happen to trigger them. If that happens then they get sent whatever it is that the trigger has been set to send (human set-up cost required again) – abandoned baskets being the commonly quoted example.
But again, by definition the triggered solution is not predicting personalisation, it has to wait for reactions on a site in order to be called into action. Consider then a true predictive email personalisation software that identifies that a particular individual is most likely to buy a specific product in the middle of next month, and it is sufficiently refined to know the colour, style and, dependent on your vertical, it might be cut, brand, price bracket – for fashion; or region, year, grape variety, chateau etc for wine etc – you can see how it works for your vertical.
A predictive personalisation email solution is still able to send that individual their unique selection, autonomously without the need of staff and now without the need of that individual coming to your site to provoke the action. It already knows, and if that customer happens to be considering wandering to a competitor, your email will have been timely received, using the data (knowledge) otherwise unused sitting in your system achieving nothing. Their email populated with the highest raking products with the greatest likelihood of being purchased by that person at the precise moment. Tomorrow those offerings would be different.
Conclusion
Beware the wide-lapelled, slick talking websites that tell you their offering is leading the way in personalisation, it just ain’t. Half of the problem is conveying the minutia of the capabilities of predictive email personalisation software. In that it has a place as well as existing solutions, but which delivers the highest ROI in ecommerce marketing. That’s more that 20x greater than everything else you employ combined. And it it does this with precision time and again, without the necessity for any staff. No wonder it merits immediate installation.
SwiftERM is a Microsoft Partner company. For a month free trial of predictive personalisation email software follow the link.
