If you've been to a marketing conference this year, you already know: 2018 is the year of personalization. It's on every trend list, every keynote slide, every vendor pitch deck. And honestly, the case is compelling.
But here's what nobody on stage wants to admit. Most companies are still pretty bad at this. Only 12% of marketers say they're "very" or "extremely" satisfied with their personalization efforts. We've declared an entire year in honor of personalization, and almost nine out of ten of us feel like we're falling short.
The best are pulling away, fast
I was catching up with a friend who works on Spotify's product team last month, and he said something that stuck with me: "We don't think of personalization as a feature. It's more like oxygen. Take it away and the whole experience suffocates."
Netflix's recommendation engine drives 75 to 80% of everything people watch on the platform and saves them around a billion dollars a year in reduced churn. About 35% of Amazon's purchases come through personalized suggestions. Spotify's Discover Weekly has become the kind of feature people evangelize without being asked.
What these companies share isn't a magic algorithm. It's that personalization is woven into their architecture, their teams, their entire way of building product.
Meanwhile, the rest of us are stuck at "Hi, {first_name}"
Here's a small, dumb thing that happened to me last week. I'd been browsing for a housewarming gift for a friend on a home decor site. Spent maybe twenty minutes looking at ceramic vases. Bought one. Done. For the next two weeks, every ad on every site I visited was ceramic vases. I already bought the vase. The moment had passed. And yet the internet was convinced I was building a vase collection. That's the state of personalization for most companies right now.
At work, the picture isn't much better. At my last role, we spent six months evaluating personalization platforms. Demos, comparison spreadsheets, meetings that could have been emails. When we finally picked a vendor and tried to plug it into our stack? Our data was a mess. Customer records lived in four systems that didn't talk to each other. We spent the first three months just cleaning data before we could personalize a single thing.
Only about 24% of marketers have a tech stack that lets them personalize across all touchpoints. And Accenture Interactive found that 48% of shoppers have left a website and bought somewhere else because the experience felt generic. That's real money walking out the door.
What the winners do differently
Teams getting real traction share a few habits. They start small, picking one high-impact use case rather than trying to personalize everything at once. They invest in data infrastructure before shopping for personalization software. They measure obsessively, running A/B tests to quantify the lift.
And the biggest one: they build cross-functional teams. Product, engineering, and marketing working on it together as a shared problem. That shift alone changes how fast you can learn and iterate.
So where does that leave us?
The year of personalization will end, as years do. The companies that made real progress weren't the ones with the flashiest tools. They were the ones who treated personalization as a product discipline, one that demands cross-functional ownership, small starts, and patience.
Personalization isn't a feature you bolt on. It's a way of building.