SEO is an efficient market

If an advantage is cheap to act on, an efficient market will automatically price it in.

Search is an efficient market. Any advantage that's cheap to act on gets priced in quickly.

The basic strategy is the same, but the execution bar gets higher. You still need to write something worth reading and get real sources to vouch for you. That advice has been public for twenty years and you can't compete it away.

The cheaper a tactic is to run, the faster it stops working.

Look, this isn't a particularly novel take, and many similar posts end with "stop chasing shortcuts and focus only on quality." But that too is a bit misguided and not what I suggest.

The market corrects itself through an adversarial loop

You do something -> your competitor sees it and does it -> the schmuck behind him does it -> Google patches it -> the slate gets wiped clean.

It's adversarial with your competitors, and it's adversarial with Google itself. We are literally trying to game an algorithm without looking like we're gaming it. And the second you can game it, everyone can game it.

Google isn't out to get you, so don't take it personally. Google has no idea you exist. Cheap tactics get crowded, crowded tactics turn into patterns, and patterns are the one thing an algorithm is actually good at finding.

I'll leave the full list of every Google update to someone else because I'm lazy. You can watch the same boom and bust play out across core updates and spam updates, year after year.

Two modern examples

Everyone is looking for shortcuts on SEO and AEO. Here are two that people are running right now:

Mass publishing. The cost of producing a page is nearing zero, so naturally people started shipping hundreds of pages per month. Google can't immediately tell what pages are valuable and well researched, and which are pure slop alone. Having a lot of pages isn't necessarily you being a bad actor (ex ebay, airbnb, crunchbase….but no your b2b saas product likely doesn't need 10K pages).

Scaled content abuse is its own line in Google's spam policies now, and it doesn't care whether a person or a script made the pages.

We see pages get indexed → de-indexed → penalized → then you’re screwed.

Self-serving listicles. You publish "10 Best [Category] Tools in 2026," put yourself at #1, and fill the rest with competitors. It ranks for "best X" commercial queries, it feeds AI Overviews and ChatGPT, and it genuinely works. It works so well that companies started trading placements with each other. I get emails from competitors asking for this constantly.

So what happens when every vendor in a category publishes the same list with slightly different ordering?

Once everybody says they're #1, "who says they're #1" tells you nothing, and the format stops being information. Backlinks obviously play a role here, but we can talk about this another time.

Around January 20, 2026, Google shipped an unconfirmed update, and Lily Ray found the pattern across 40+ sites: drops of roughly 29% to 42%, spreading well past the folder the listicles lived in. It gets worse in AI search, where Google will cite your listicle but recommend a competitor inside the same answer, which she measured at about 69% across 100 B2B queries.

Note: 100 queries is a super tiny sample size. We'll run our own experiment at Maintouch in time.

Okay, so what’s the actual lesson then?

The lesson isn't that these tactics don't work. They obviously work, and people are making serious cash executing on them.

The lesson is that the same tactic can be incredibly good and then incredibly bad. It can put you in the AI answer, hand the recommendation to your competitor, or get your site nuked.

With mass publishing, volume is at least a lever you control. You're sizing a bet in a market where you can't model risk well. The machine learning engineer in me is frustrated by this.

So run the tactic and take the short-term win. Learn what made it work, because underneath most of these there's a real insight about buyer intent that outlives the tactic itself. Just don't let it become your foundation. When it unwinds (and it will unwind) you want it to cost you a quarter, not your domain.

Don't let AI do 100% of your SEO

Rich, coming from me. Say what you want.

Not for purity reasons, I don't care about purity. For market reasons. Anything your agent does off a generic prompt, your competitor's agent does off the same generic prompt for the same $20. That is, by definition, no edge. A general model running a general prompt produces the most commodity object on the internet, at volume, which is exactly the thing the last four years of updates were built to find.

Garbage in = garbage out. The stuff that can't be replaced is what's valuable: context from sales calls, support tickets, proprietary data, actual subject matter expertise, off-page work that needs a human to care.

(FYI: most of this is literally just E-E-A-T. look it up)

AI search is NOT efficiently priced (yet)

Google is efficient because Google has the perfect feedback engine. Every click, every bounce back to the results, every refinement, for twenty some odd years.

LLMs don't have this loop yet, and admittedly, they may never have it. The way they train these models is based on online behavior/responses, not off-platform behavior.

The window on AI search is wider, and, at least for now, it's staying open longer. There's more room to exploit AI search right now than there is to exploit Google.

But the loop is being built while you read this. Every one of these systems is indeed trying to build their own version of this. We see this with ChatGPT showing links and map listings more prominently. Additionally, because LLMs can reason on top of context, they'll be able to see through a lot of self-promo stuff in time.

Just know that it's a window.

Why this matters

Everything above is the difference between a strategy and a trade. If a tactic only works in the short term, it's a trade. Most of what gets sold as SEO and AEO strategy is really a stack of trades, and the strategy is the reason any of them worked in the first place.

So run the trades, but size them like trades. The strategy underneath is the same one it's always been: own the things that are expensive to copy.

That's the entire edge. Expensive is the one thing an efficient market can't compete away, because your competitor could copy it and won't.

Anyway, if you want to jam on strategy and trades, we built Maintouch to run the machinery around all of this. Get my (current) playbook ⟼ maintouch.com/demo