Put your prices on the page

Ask any agency what to do about AI visibility and you will hear about structure. Headings. Structured data. Clean formatting. Bullet points. Tables. Make the content easy to skim, easy for software to read, neatly arranged.

In May 2026 three researchers set out to test that properly, and the result is the most uncomfortable finding in this field for the people selling it.

It is the best-built study on this question that exists. 252,000 trials across six AI models. 1,440 comparisons, drawn from a hundred real product pages with the names taken off, across fifty categories. Each comparison sets two pages side by side that differ in exactly one thing: the brands removed, so nothing wins on fame alone; the lengths kept almost equal; and the position in the list swapped back and forth, so that being first could not skew the result.

Eighteen factors tested. Four came out as gatekeepers, significant across all six models:

How much the page is actually about the subject    cited enormously more often
Being at the top of the list                      cited enormously more often
Carrying a recent date rather than an old one     cited at least 14 times more often
Saying what something costs rather than not       cited at least 6 times more often

Three of those four are unsurprising. Being about the subject matters. Being first matters. Not being seven years out of date matters.

The fourth is the one nobody talks about. Whether the page says what something costs sits right up at the top, next to being about the subject and being first. We have not found a single published guide to AI visibility that recommends it.

Now the other end of the same table, the weakest tier, where the result held in three models or fewer:

Neatly formatted rather than dense text     from slightly worse to slightly better

Read that range carefully: it starts below break-even. In some models, tidying up the page made citation slightly less likely. The single most-recommended tactic in this entire industry, properly measured, cannot be told apart from doing nothing.

The caveats are large and you should weigh them. It was published before other researchers had checked it. The test pages were written by one AI model and graded by others: the three authors reread three hundred of them by hand, which is diligence, but it is still the same family marking its own homework. The brand names were stripped out, so the study says nothing about how a published price fares against a famous name in the real world. And the authors are candid about the ceiling of their method: they compare two pages at a time, while a real system pulls five or ten together, where these things play off each other in ways a two-page comparison cannot show.

So this is one unreplicated result. Treat it as a hypothesis worth testing rather than a law.

Except it is not quite alone. Two independent lines of evidence point the same way.

Kurt Fischman analyzed 730 citations across 1,006 pages and found that, at equal rank on Google, the descriptive code behind a page changed nothing. One exception: pages where the product and review fields had actually been filled in, with real prices and real specifications, were cited substantially more. The code was not doing the work. The facts inside it were.

And when Advanced Web Ranking sat down and read 112 paragraphs by hand, comparing the cited ones against the ignored ones, they recorded a finding worth memorizing: among the paragraphs nobody cited, not one contained a new claim or a hard number. Not few. None.

Three different methods, one direction. Specificity gets quoted. Presentation doesn’t.

There is a plain mechanical reason for this, and it has nothing to do with optimization. A model building an answer to “how much does X cost in Y” needs a number. If your page says “contact us for a quote”, there is nothing to lift: you have written a page that cannot be used to answer the question your customer asked. A competitor who wrote “$400–$900 for a typical three-bedroom, same-week availability” has handed the model a sentence, and the model will use the sentence it has.

This is also the item clients resist most, and the objection is always the same: our pricing depends on the job. It usually does. But “it depends” and “we won’t say anything” are two different positions, and a range with the reasons named, $400 to $900 depending on roof pitch and access, is both true and quotable. The businesses that refuse to publish any number are not protecting their margin. They are removing themselves from the set of answers a model is able to give.