AI search changed the reader: first you convince the filter, the human comes second
I hardly even open the first page of Google any more, never mind the second one.
There is an indecent amount of content
You can curse the neural networks or you can bless them: it makes no difference, they have already changed the market and that mince doesn't go back through the grinder. The entry threshold for production has fallen to cents, and everything that previously simply wasn't born out of laziness and lack of budget has poured into the feed.
Attention doesn't get manufactured by pressing a button
The cost of a touchpoint, meanwhile, is going up everywhere. And attention, unlike content, doesn't get stamped out in series: the person has the same twenty-four hours, the same tiredness and the same wish to close the tab and go and live in the forest to escape the overload. They had already been squeezed dry before this whole story with AI, and now a great deal more people have shown up at the exhausted well.
The industry answers with volume
If the touchpoint gets more expensive, come on then, let's generate five times more content, set up content factories on a conveyor and plug every crack with the same neural slop. That is already in poor taste. And generated slop is impossible to tell apart from the real thing, which a lot of people exploit rather skilfully.
The most interesting part happens one floor up
A new layer has settled on top of all the content: neural networks that filter it instead of the user. The person leafs through the results less and less and collects opinions crumb by crumb: they ask the model and receive a ready answer with the sources already chosen for them. I notice it in myself: every month I google less and ask more. That is what zero click search means in practice - the answer arrives and the click never happens. AI Overviews sit at the top of the page, an AI search assistant answers instead of the page, and recommendation algorithms were the first bell many years ago. Now between you and your audience there is an intermediary that decides on its own whether you even deserve to be shown.
The intermediary has different rules
You are no longer fighting for the person to click on your headline, but for the machine to consider you a sensible source and retell you in its own words. And the rules here are very different: the intermediary absolutely does not care about clickbait or about how neatly you intrigue in the first line. It looks at whether there is structure, facts and specificity underneath the words, and whether there are traces of your presence in the places it pulls the data from: reviews, forums, Reddit, opinions, live discussions.
It's a funny turn: the market spent years polishing the art of manipulating a living person, and in the end the main reader turned out to be a gadget on which those tricks have no effect at all.
The practical question everyone asks next - how to rank in ChatGPT, what ChatGPT SEO even is - has its own long answer, and it is in Optimizing for a reader that isn't human. This piece is about why the question appeared at all.
What this changes in practice
If the decision about what to show the person is increasingly taken by a machine, then the touchpoints have to be systematic and continuous - that resembles a process more than a campaign, and not a heroic manual feat once a quarter, when somebody finally has hands free.
That is why the product stepped further than research, into the automation of touchpoints and traffic: SEO Traffic Growth, so that you get found both in the search engine and in the answers of the neural networks, and Social Traffic Growth, a distribution that isn't rented from anybody. For now it's a beta: it's being run in with one corporate client, live, on their real tasks and their own stumbles. I'm not going to pretend I have all the answers: I don't have them, but the direction feels right.
Conclusion
The substance hasn't changed, the addressee has. First you convince the filter, and only afterwards the person.
