Twenty years ago, in early 1999, the HBR published a pioneering article on the future of personalization titled “Is your company ready for one-to-one marketing?”
Reading it is a flashback to the heady days of the dot-com era, with Amazon as case study number one. The article states:
“Relationship marketing is grounded in the idea of establishing a learning relationship with each customer, starting with your most valuable ones. Think of a learning relationship as one that gets smarter with each interaction. The customer tells you of some need, and you customize your product or service to meet it. Every interaction and modification improves your ability to fit your product to this particular customer.”
Today, the customer no longer has to “tell you of some need”, the machine learning personalization algorithms can figure that out. More than 80% of the TV shows and movies people watch on Netflix are discovered through the Netflix recommendation platform. When we think we are choosing what to watch on Netflix, we’re really just choosing from decisions made by their algorithm that increasingly and accurately predicts what we’ll like.
That approach to personalization is starting to spread to marketers everywhere. Evergage recently published a “2019 Trends in Personalization” report and found that usage among marketers of machine learning/algorithmic personalization is up from 26% last year to 40% this year.
They also found that 85% of marketers believe that their customers and prospects expect a personalized experience, but only 32% believe marketers are currently getting personalization right. Among marketers who use machine learning, 78% described personalization’s impact on customer relationships as “strong” or “extremely strong.”
As with so much marketing technology, we’re in an awkward adolescent period, where technology is increasingly making this 20-year old promise a reality. And yet, we’re working through the myriad unintended consequences of machine learning, and continually figuring out the line between cool and creepy.
Here are a few related cartoons I’ve drawn over the years:
“Personalization” November 2014
“The Future of Advertising”, August 2013
“Marketing with Personal Data” May 2014
Bill Stiles says
Personalization often seems no more personal than a billboard. Most consumers have eclectic tastes that the algorithms overlook. The narrow recommendations become a disservice, suggesting products I’ve already purchased or content that leads to boredom. In both cases we learn to mistrust the recommendations and ignore them.
There is no shortcut to effective marketing.
Ori Pomerantz says
I want content that *I* find interesting. It makes sense to show me things that people with similar tastes enjoyed. That’s personalization that’s good for me as the customer.
OTOH, it doesn’t make sense to show me ten ads for the same product in an hour. If the first nine iterations didn’t work, the tenth won’t either. It also doesn’t make sense to show me five articles about the Orville being renewed – one would have been enough.
Jason says
One Achilles heel of personalization is that you can deduce what a person might well like – but you can’t know whether they’ve then seen it. Thus your well-intentioned algorithmic suggestion merely becomes another minuscule irritation that further debases and demeans advertising and advertisers.
Allen Roberts says
The cause and effect chain gets worse as you consider the consequences of AI creating the personalisation without adult (human adult) review.
This terrific post by Josh Bernoff makes a scary point. https://tinyurl.com/y379lpsc Perhaps worth a follow up cartoon?
Keith says
I agree with Bill above. Showing me ad’s for thing’s I’ve already purchased is more often than not counter-intuitive. I’ve purchased and moved on, in most cases I wouldn’t need that product again. Also as you have previously noted, we don’t expect a store to follow us home or to another store, once we have left that store, we have moved on. That’s the same here, often when we have purchased something, we’ve then moved on, showing adds of a product I brought say, while I’m on holiday, when I get back to work, the context and emotions are all wrong, and more often than not, that makes us switch off to the adds.
Often purchases can be tied to a time, emotion, reason, showing an add outside of the perimeters is a total waste of the person’s time.
And the overload we get of this now, honestly those adds might as well be black holes, they simply have no effect.