Personalization and Machine Learning –Are We There Yet?

Senior CX Designer
Valtech DK

fevereiro 06, 2018

Everyone has come to expect relevant content when they’re in the digital arena. Everyone’s talking about delivering just that. Personalization is everywhere. But the truth is that no person is smart enough to deliver it. And no person ever will be. If we are to deliver on the promise of personalization, machine learning and the democratization of it, is essential.

First, What’s all this personalization-fuzz about?

Many proclaimed that Personalization was the marketing buzzword of 2017 – true or not, it surely didn’t go unnoticed by anyone who’s even remotely associated with someone in digital marketing or sales. And with good reason. We’re all fed up as consumers, with irrelevant content being thrown our way left and right. We want the right content at the right time and that’s what personalization is supposed to provide us with. A new era was somewhat proclaimed, where we, as consumers and digital beings, were drawn in by the idea of a future of nonsense-free content. Every CMO could almost physically feel the booming of their business, with the prospect of being able to provide every user or customer with a digital journey tailored just for them.

The Media industry has definitely been a first-mover when it comes to actual personalization, and on their way, they’ve demonstrated exactly what will happen to those who don’t keep up. Meanwhile, other industries are merely scraping the surface and in most cases they’re not actually personalizing at this point.

When people say they’re personalizing – they’re probably doing advanced handheld segmentation

When you move around on most digital platforms, you’re treated as part of a segment. Most often, you were put in a box, by another person, who programmed their site to register you as part of group X, when you did Y and Z. You and everyone else who had similar digital moves. As a result, the uniqueness you are treated with, is based on that other person’s ability to analyse your behaviour and relate it to other needs. This does, in most cases, provide a much, much (!), better experience for consumers, however, it’s not personal, it is certainly not flawless, and it could definitely be much better.

So why aren’t we personalizing yet? – that’s perhaps your number one question right now – Because it’s a very poor business case for companies, as they would need to hire hundreds of employees to be able to address the unique needs of individual customers.

That’s why actual personalization is completely reliant on the development of machine learning for personalization.

How machine learning will lead the way to actual personalization

One of the major limitations of personalization efforts today, is the ability of the human brain to understand and manage larger amounts of data. Compared to what we’ve managed to teach machines, we’re mere infants trying to juggle and manage data that’s far beyond our reach and capabilities. With our limited capabilities, it would require a small army of employees, for any business to do actual personalization. Luckily, we’re smart enough to programme machines to manage enormous amounts of data for us.

By using the amazing capabilities offered to us by machines, we can move from hypothesis-based segmentation, tagging and scoring, onto recognition of the behaviour and paths taken by any individual similar to previous paths and behaviours by other individuals, and create our recommendations and further movement on that. This way, in time, the individuals engaging digitally with us, will be guided through our content based on actual behavioural patterns rather than paths predicted by a human brain – with all its limitations.

A selected few are already using machine learning – but what about the rest of us?

Some of the very biggest of corporations are going full speed ahead and provide personalization based on machine learning. However, we’re talking about a very limited number, who are only capable of doing so, because of their massive size, unique setup and capabilities far beyond the reach of the vast majority of companies. And even in these very limited cases, they are not taking full advantage of the immense possibilities.

Naturally, like any other development in history, a democratization of this technology is imminent. Consumers are demanding it, marketers are longing for it and we can’t wait to help deliver it.