What do Google and ChatGPT have in common?

The first index entry was a hand

Roughly forty thousand years ago, in a cave on Sulawesi, somebody pressed a palm to the wall and blew powdered pigment around it. What was left is an outline, a negative space in the shape of a person. It is close to the oldest message we have from anyone, and it says one thing.

I was here.

That is not decoration. It is a record, made so a stranger arriving later could retrieve it. Every technology since is a variation on the same move.

A note on that number, since the second half of this essay is about being careful with numbers. These dates measure the mineral crust that grew over the paint, so they are minimums, and the method is contested. At one Spanish site, five samples taken within five centimetres of the same hand returned ages of 14,700, 23,100, 35,300, 55,200 and 66,700 years.

Only the oldest gets quoted. Hold that thought too.

The library was the first time knowledge outlived the knower

Alexandria did not invent writing. It invented the pile. Put enough scrolls in one room and the room knows more than any person in it, and keeps knowing after they die.

Ancient sources put the collection anywhere from forty thousand scrolls to seven hundred thousand. The classicist Roger Bagnall argued in 2002 that essentially all of those numbers are fantasy. Even the founding myth of information storage is a claim nobody checked.

IV. The mirror

In 1998, nobody knew what to do with the box either

Marketers bought directory submissions. They stuffed the meta keywords tag until it read like a ransom note. They traded links in rings with strangers.

Most of that was dead within five years. The people who won were the ones who worked out early that a link was a vote, and behaved accordingly, while everyone else treated links as decoration.

V. How people actually decide

Step away from marketing for a second

Think about the last thing you bought that cost more than a dinner. You did not build a spreadsheet of ten options and score them on weighted criteria. Nobody has ever done that.

You narrowed to two or three almost immediately, on grounds you would struggle to defend, and then you looked for permission to pick the one you already wanted.

In person, a conversation comes with scaffolding

You never notice it because it is free. The other person asks a follow-up you did not think to ask. You hear them hesitate before the word "probably" and you weight it correctly. You watch their face lose interest and you change tack.

Most of all, you get permission to say "wait, what?" without feeling stupid. That single affordance is doing enormous work, and no interface offered it for twenty-four years.

VI. Forward

Live is the next scaffold

Voice that interrupts you. Screen share where the thing watches you work. Agents that ask a clarifying question before doing the task instead of after ruining it.

It is the same move again, and it points one direction. The interface keeps bending toward how people already behaved. Technology has been catching up to us since the cave wall. It was never the other way around.

VII. The playbook · One

Baseline before you touch anything

Write twenty-five prompts a real buyer would type. Not keywords. Sentences, with the mess left in. Run every one through ChatGPT, Perplexity, Google AI Mode and Claude, and log three things: were you named, were you cited, and who got named instead.

That grid is your GA4 for LLMs. Most brands have never once looked at it, which is remarkable given how much they spend guessing.

Do it this week: twenty-five prompts, four engines, one spreadsheet. It takes an afternoon and it will change what you argue about in your next meeting.

If an afternoon is more than you have right now, run the free AI visibility audit and it will do a first pass on whether the engines name you at all.

Three

Earn agreement somewhere other than your own website

Your site is one vote. It is a vote the model knows you control, which is exactly why it is worth less than you want it to be.

Ahrefs tested this across nearly ten thousand AI answers in mid-2026. For a brand that already had third-party coverage, 94% of new mentions came from content it did not own. Only 6% came from its own pages.

Now the complication, because this is where most GEO advice lies to you. For a brand with no third-party coverage, the numbers flipped: 82% came from its own pages. And the largest study published, 11.84 billion citations analysed by Profound, found company websites make up about 57% of all AI citations.

So "Reddit runs the models" is true at the top of the distribution and false across the whole of it. Your own site still matters. It stops being sufficient the moment other people start talking about your category.

Do it this week: pick the three surfaces where your buyers already argue with each other, and go be genuinely useful on them under your own name.

One warning. Google's spam policies were updated in May 2026 to name "attempting to manipulate generative AI responses" as spam. There is a real line between earning coverage and manufacturing it, and anyone who bought links in 2005 knows exactly where it is.

Five

Write to be lifted

There is peer-reviewed evidence for this, which is rarer in this field than the confidence levels suggest. The KDD 2024 paper that named generative engine optimization tested nine ways of editing a page and measured visibility across ten thousand queries.

Adding quotations lifted visibility 42.6%. Adding statistics lifted it 32.8%. Citing sources lifted it 27.7%. Keyword stuffing lowered it 8.8%, which should settle an argument some people are still having.

Do it this week: put a self-contained summary above your intro that answers the question completely on its own, turn your headings into the questions people actually ask, and add a comparison table to anything with commercial intent. Models lift tables constantly.

Seven

Entity hygiene, and I do mean hygiene

Point your organisation markup at Wikipedia, Wikidata and your verified profiles. Keep one consistent identifier across your graph. Make your brand facts identical everywhere they appear, because a model resolving you to two entities is a model that trusts neither.

This is disambiguation. It is not a growth lever, and anybody selling it as one is about to become the subject of the next section.

The AI Mode and ChatGPT effects came in at plus 2.4% and plus 2.2%, neither significant. Four separate tests agreed, and the study found no vertical, no authority level and no schema type where it helped.

A second study, on SSRN in February 2026, is sharper. Attribute-rich schema was cited 61.7% of the time. No schema at all scored 59.8%, a difference that is not statistically significant. Generic Article and Organization markup, the kind most sites actually ship, scored 41.6%. Worse than nothing.

The strongest argument against me

In April 2026, AirOps and Kevin Indig analysed 16,851 queries and found pages with JSON-LD cited 38.5% of the time versus 32.0% without. They controlled for word count, headings and domain authority, and called it an independent signal.

Here is why I think they are measuring a shadow. They did not control for retrieval rank, which was their own biggest finding: position one cited 58.4%, position ten 14.2%. That effect is four times the schema gap, and pages that bother with structured data are systematically the pages that rank. The Ahrefs design holds the page constant and measures before and after, which is the stronger instrument for a causal question.

What to do: ship schema for entity hygiene and rich results, which are real. Do not ship it expecting citations, and do not let anyone bill you for it as AI visibility work.

Lily Ray tested 100 commercial queries across AI Overviews in mid-2026. When a brand's own "best of" page got cited, that brand was left out of the actual recommendation 69% of the time. The model read their page, took the list, and recommended a competitor from it.

The Ahrefs experiment found the same pattern independently, at 43%. Two different teams, two methods, same uncomfortable answer.

So a dashboard showing your citations climbing can be showing you a machine that is politely using your research to sell somebody else's product. Measure whether you are named in the answer, not whether you appear in the footnotes.

That is the specific thing the free audit checks: not whether you were quoted, but whether you were recommended.

Which brings us back to the wall

Forty thousand years ago somebody put a hand on rock and blew pigment around it, and the message was I was here. Every system since has been a way of asking that question on somebody else's behalf and getting an answer back.

Google answered it by counting who pointed at you. The models answer it by counting who agrees about you. Neither one has ever had an opinion. They just tally.

If you want to know how many walls currently do, the free audit takes about a minute.