For the past decade, we’ve been sold a seductive story about artificial intelligence. The promise was simple: AI would let us know our customers better than they know themselves. We’d learn their favorite color, their pet’s name, the exact moment they scroll through Instagram. Then we’d serve them the “perfect” ad.
I’ve watched agency after agency build their entire pitch around this concept. I’ve seen the slide decks. I’ve heard the buzzwords. And I’ve watched the results plateau.
Here’s the uncomfortable truth: hyper-personalization is hitting a wall. And most marketers don’t want to admit it.
The Uncanny Valley of Marketing
Let’s be honest with ourselves. When you buy a pair of shoes and spend the next two weeks seeing ads for those exact shoes, do you feel understood? Or do you feel watched?
That feeling is what I call the “uncanny valley of marketing.” We’ve trained consumers to expect relevance, but we’ve overcorrected into creepiness. Reminding someone of a past action isn’t clever. It’s lazy. Serving an ad for a product they just looked at isn’t engagement. It’s stalking.
We’ve all become numb to the retargeted toaster.
True engagement doesn’t work that way. It never has.
A Different Way to Think About AI
Here’s what I’ve learned after years of running campaigns across every major platform. The most powerful application of AI isn’t personalization. It’s something I call anticipatory empathy.
Instead of asking “what did this customer buy?” ask “what does this customer need to stop worrying about?”
That shift changes everything.
On TikTok and Instagram
High-emotion platforms are terrible for hard-sell personalization. Users are there for escape. They aren’t shopping. They’re scrolling.
The real opportunity is latent intent prediction. Instead of targeting someone who searched for “best hiking boots,” train your models to identify users posting about “sore feet” or “getting back in shape.” The AI identifies the problem before the user types a single search query.
You’re not interrupting their scroll with a product. You’re showing up with a solution to something they’re already feeling. That’s a conversation starter. Not an interruption.
On Google and YouTube
Search ads are the bastion of high-intent, low-empathy marketing. Someone types “cheap flights to Paris.” You serve them a cheap flight ad. Yawn.
The smarter play is friction de-escalation. Analyze the syntax of the search query. Is the user asking a question? Using words like “troubleshoot” or “urgent”? That signals anxiety, not intent.
For a software company, a search for “how to uninstall [competitor’s software]” is pure gold. Serve them an ad not about your features, but about a seamless migration guide. You’re engaging with their fear of switching, not their desire for features.
On Pinterest
Pinterest is aspiration, not desperation. Users dream and plan. Most personalization looks backward at what someone pinned. Anticipatory empathy looks at the gap between their boards.
A user has a board for “Dream Kitchens” and a board for “Budget Renovations.” The AI spots the tension. Don’t just show them a faucet. Offer them a guide titled “How to get a $50k look for $15k.” You’re selling the bridge between where they are and where they want to be.
Why Most Agencies Won’t Do This
Because it’s harder. And because their incentive model is built on volume.
Most agencies are measured on:
- Impressions served
- Clicks generated
- Cost per acquisition
Anticipatory empathy requires a different set of metrics:
- Reduction in downstream friction
- Fewer returns and support calls
- Shorter sales cycles
- Higher customer lifetime value
These metrics are harder to track. They require a different relationship with data. They require an agency that’s actually aligned with your business outcomes, not just your ad spend.
Where to Start
If you want to apply this thinking, here’s a simple framework:
- Stop asking what your customer wants to buy. That’s a waste of processing power.
- Ask what your customer needs to stop worrying about. Identify the friction points in their life or business.
- Show up with a solution before they ask for it. Not a product. A pathway.
- Measure what actually matters. Look for signals of reduced confusion and increased trust.
The future of engagement doesn’t belong to the loudest voice. It belongs to the one who arrives first, with the right answer, to a question the customer hasn’t yet asked.
That’s not personalization. That’s empathy. And it changes everything.
