Tinuiti, the largest independent performance marketing agency in the U.S., just published a smart, thorough guide arguing that AI search has changed the rules of growth and that you need a whole new operating model to win.
They’re right about a lot of it. But I think the headline is wrong.
AI did not fundamentally change search. It reset the tactics, not the game.
Here’s what I mean. Five years ago, Google took the best technology on earth and pointed it at one job: find the most valuable, relevant websites and put them in front of people. The sites that ranked highest were, broadly, the ones that were actually useful.
Today Claude and ChatGPT are doing the exact same job. They read the web and try to surface the businesses and pages that are genuinely valuable to the person asking. Different machine, same mission.
So if you want to rank number one in AI, the answer is boring and it hasn’t changed. Be a genuinely valuable business, with a website people actually enjoy using. That’s the 80. Everything in a 6,000-word AEO guide is the 20.
The tricks aren’t useless. They’re just temporary.
Remember early Google? People stuffed pages with keywords. White text on a white background, the same word typed hundreds of times where visitors couldn’t see it, because early Google mostly counted how often a term showed up on the page. It worked. Until everyone did it. Google patched it, and overnight the trick didn’t just stop working, it started hurting you.
That story is the whole story. Every clever exploit of a ranking system has a shelf life that ends the moment it gets popular. Schema gymnastics, entity-stuffing, prompt-baiting, whatever this quarter’s AEO hack is, the models will re-weight it on the next update. And AI models update a lot faster than Google ever did.
So over-optimizing for the current tricks is a treadmill. You do the work, the model shifts, you do it again. Meanwhile the one thing that never gets patched is being worth recommending.
To be fair to Tinuiti
Their guide actually says this in the fine print. They call E-E-A-T and genuinely useful, “non-commodity” content the core signals, and they note that the fundamentals don’t change, only the surface area does. I agree completely. I’d just flip the emphasis. The fundamentals are the point. The surface area is the footnote.
Here’s the practical version
- Build something people are glad they found. If your website is genuinely helpful, clear, and worth someone’s time, you’re already aligned with what every model is trying to reward.
- Learn the current tactics, but hold them loosely. Clean structure, clear answers, solid technical hygiene — fine, do all of it. Just don’t bet your business on any single trick surviving the next model.
- Ignore anyone selling a permanent hack. It doesn’t exist. It never did with Google, and it won’t with AI.
The businesses that panicked over every Google algorithm update spent a decade exhausted and behind. The ones who just kept being genuinely good compounded. Same movie, new projector.
AI didn’t change the fundamentals of search. It just gave everyone a fresh reason to relearn them.
If you want help building the kind of business AI actually wants to recommend, that’s what we do. Come say hi at north.digital.
