Google Is Lying to You About AI Search (Here's the Truth)

You said that Google is gaslighting SEOs, right? We've heard this almost every time they come out with some SEO guidelines.

What did they get wrong specifically in this new AI SEO guideline?

Yeah, so there's a bunch of information that is sort of a difference between obfuscation and direct saying something is not correct, right? Those are kind of two separate things.

The obfuscation side of things, that's where they're saying you shouldn't use schema, you don't need to put markdown versions of your content, you don't need to do those kind of things.

However, if you actually look at their own products, they're doing all of those things for their own websites.

Even the blog that they put those guides on, they're still doing all of that.

When it comes to direct saying, you don't need to do this because this is not how things work.

Things like extractability and chunkability and sentiment and things like that, they 100% do matter.

And Google's current AI guide, the engineers going on this tour, all of those kind of things, they're all pointing against that.

and a lot of the models especially if you're looking at ao overviews there is exact chunks of content that they're going to extract and summarize into the ao views and then cite hopefully cite you back for being the source of that original chunk so a lot of those bits where they're saying chunks don't matter extractability doesn't matter these things don't matter they 100 do and a lot of the guidance isn't from the people who are actually creating those models, who are working day to day to make those models better.

It's the same people who have created the documentation, the policy for Google's own departments for many, many years previously.

And again, we've seen throughout those 10, 20 years, if you look at the reality or you look at leaks or you look at DOJ testimony, a lot of what they're saying on those policy documents and the documentation is factually incorrect versus reality.

So the question then is why, right? Why do they keep doing this, Charles? What's the incentive?

So the incentive originally for SEO is that they don't want to have people using certain techniques, using certain massive spamming campaigns, all that kind of stuff, because it absorbs some of their resources.

And if they can use... few resources, propaganda and PR and conferences and SEO influences and things to spread that messaging, which is very, very small amounts of resourcing to do that versus having to create systems to actually counteract these things.

In reality, those things are very expensive, take massive amounts of processing power, especially right now.

They're trying to put all of that processing power into AI, into new iterations of their models, into new intelligence, into new tools and suites, all these kind of things.

I think that over the next several years, in terms of the AI version of that policy of that documentation, it's more to do with how they're trying to represent themselves and how they're trying to represent their models.

And it's more of a corporate approach.

presentation to the investors and to the digital PR world and to the kind of more corporate side of that world to show that there is something that they're trying to help you with, but realistically how helpful that guide is is up for debate.

I've always thought it's super ironic too, right? Because SEO at its core has always been black hat, you could argue, right? Your whole idea here is you're trying to manipulate an algorithm or manipulate algorithms just outside of Google to try to show up in the top result.

It's no different here in the AI search world.

I just find that super ironic.

But let's talk about rank tracking here because attribution has always been the devil of marketing and SEO.