The question
Everyone in this field gives advice about where to put things on a page so an AI will use them. Put it in schema. Put it in a table. Keep it out of JavaScript. Almost none of that advice is tested, and the providers themselves document very little of it.
So: if the same page carries the same kind of content in seven different places, which of those places can an AI actually retrieve from, and how long does each take?
Predictions, on record, before any data
Written 22 August 2026, before a single page was published. They are here so you can check them against the results, including the ones I get wrong.
- Plain body paragraph text resolves first, inside 14 days
- Image alt text and structured data never resolve on any engine
- JavaScript injected content takes over a month on at least three of the five engines
- The llms.txt entry does nothing at all
If all four are wrong, the table gets published anyway. That is the entire point of writing them down first.
Method
Each test page carries seven unique nonsense strings, called markers. A marker is a string like vurpalis7742 that exists nowhere else on the internet. If an AI engine can tell you what one is, the only possible source is the page it sits on. That is what makes the result unfakeable.
Every marker was checked on two search engines before publication to confirm it appeared nowhere. Each marker sits in exactly one place, and all seven placements live on the same page, so every page level factor is identical between them. Any difference in retrieval can only come from the placement.
| Arm | Placement | Present in HTML source |
|---|---|---|
| A | Plain body paragraph, visible | Yes |
| B | H2 heading | Yes |
| C | Table cell | Yes |
| D | Image alt attribute only | Yes, as an attribute |
| E | JSON-LD structured data only, invisible on the page | Yes, inside a script tag |
| F | Inside a collapsed block, hidden until opened | Yes |
| G | Injected by external JavaScript after load | No |
Two further arms ride along, testing things that cannot live inside a page: a marker that exists only in the site's llms.txt file, and a marker in each PDF edition of the test pages. The PDF markers are separate from their source page's markers, because a marker that exists in two places tests nothing.
Each engine is asked what each marker is, and the date of the first correct answer is recorded. Browsing on and browsing off are logged separately, because they measure different things: one asks what the search index behind the engine holds, the other asks what reached the model.
The pages
Three pages, each carrying its own set of seven markers. They are real reference pages, written to be useful on their own, because thirteen pages of scaffolding would be a worse thing to put on a website than three pages worth reading.
- Every AI crawler user agent, and what blocking each one actually does
- How a page reaches an AI answer, and where it usually breaks
- What each AI provider actually documents about how it picks sources
Three rather than one, because if a page never gets ingested at all, its markers all read as silence and you learn nothing. Three gives three chances at a page that worked, and lets the results say whether the arms behaved consistently.
The marker values are not listed anywhere as a set, and never will be. Each page names only its own body text marker. Publishing the full list would put every marker in a second location and void the experiment. If you find a marker quoted somewhere other than its own page, that arm is contaminated, and it will be reported as contaminated rather than dropped.
Controls
Positive control. The body paragraph marker on each page. If it resolves, that page was ingested, and the silence of the other six means something. If it never resolves, that page produced nothing and its results are void rather than negative.
Null controls. Three markers were generated and published nowhere at all. They are asked at every checkpoint alongside the real ones. If any engine confidently answers one of them, the engine is inventing answers, and the whole round is void. Without this, a study like this can mistake confabulation for retrieval.
Schedule
Checkpoints at day 1, 3, 7, 14, 21, 30, 45 and 60. Routine checkpoints query ChatGPT, Perplexity and Google AI Mode with browsing on. Milestone checkpoints at day 14, 30 and 60 add Claude and Gemini, and run browsing off as well.
Once a marker resolves, its date is recorded and it stops being queried.
Results
The study started on 22 August 2026 and closes in late October. Results will be published on this page, in a table, as text rather than an image, including every arm that turns out to be a dead end and every prediction above that turns out to be wrong.
What this will not prove
Worth stating now rather than after the numbers arrive and start looking more impressive than they are.
One domain, three pages, one time window. Small enough that the right word for the output is a signal, not a law. It cannot separate anything true of this site or this page template from anything true of the placement itself.
All markers sit in the lower half of their page, in the same section, which tightens the comparison between them and narrows the claim: the results describe placement within a page's lower half, not placement anywhere on it.
The pages disclose that they are part of a study. That cannot be removed without publishing unexplained gibberish, and it might plausibly make an engine treat the content as low value. Unmeasurable from outside, and stated here rather than buried.
Where the pages were manually submitted for indexing, that is recorded, and it means the absolute timings describe retrieval given a nudge rather than natural discovery. It does not affect the comparison between arms, since all three pages get identical treatment.