StayStrategy
An AI visibility practice for boutique hotels, short-term rentals and tour operators. I built it to test an argument I had made in writing, ran it across dozens of businesses, and have published the whole method free.
What it is
StayStrategy is an AI visibility practice for small hospitality operators. It puts a business’s digital footprint in order so that AI assistants name it when a guest asks where to stay. The method is published in full and free at staystrategy.co.
A guest asks an assistant where to stay and gets three or four names back. There is no second page to be on. Whether a property is in that answer depends on what a model can read about it, and for most small operators that material is scattered, inconsistent, and written for nobody in particular.
StayStrategy fixed that, one business at a time. I have since exited the hands-on work, and everything the engagement used to deliver is now written down and given away.
It began as a test of my own argument.
I had written that a model now assembles an answer about an operator from whatever material exists, and that being inside that answer is what visibility has come to mean. That claim is testable. Starting a business to act on it put real money behind the position and gave me somewhere to find out whether it held.
The most honest place to test it was staystrategy.co itself: a new site, no press coverage, no budget, nothing working in its favor except the method.
What happened
Bing began citing the site in AI answers within months of the footprint work.
Bing AI search citations, staystrategy.co
Daily citations of the site in AI answers, January to July 2026
Source: Bing Webmaster Tools for staystrategy.co. Bar heights show the relative monthly level, not exact daily counts. Citations are a visibility measure, not revenue.
Google moved in the same month
One system on its own is an anecdote. Search Console for the same site over the same period shows the same inflection, and the two systems are independent of each other. Two of them stepping up together in the month the footprint was put in order is the mechanism doing exactly what the argument says it does.
Google Search Console, staystrategy.co
1 May to 28 July 2026
Source: Google Search Console for staystrategy.co. Totals are exact. The curve traces the shape of the daily impressions series and is not a point-by-point plot. Impressions are a visibility measure, not revenue.
The paid campaign returned 4.6 times its cost
Visibility is one half of the question and whether any of it sells is the other, so I ran a paid campaign against the same offer.
Save the Summer, paid campaign
Dozens of small operators engaged across the run
The first paid campaign run against this offer. Stated as ratios; the underlying spend and revenue are not published.
The payback landed immediately instead of somewhere out in the life of a retainer, which is the shape the model predicts for an offer priced to be bought once and sold through a cold channel.
What the work taught me
An answer engine deciding whether to recommend a property reads several sources at once and notices when they disagree. Getting a name, an address and a description to match across every surface used to be tidiness. It now decides whether a business is recommended at all.
Good agencies and good freelancers do not want low-spend clients, and that is arithmetic. Small gigs are a way to work up to bigger ones, so the incentive runs toward graduating away from exactly the businesses that need the most help.
I feel this at the mechanic. I know nothing about cars, every quote sounds higher than it should be, and I have no way to tell afterwards whether the thing was fixed properly. Most small operators are at the mechanic when they buy marketing, which is why publishing the method does more for them than selling it did.
Once a footprint is correct, each month asks very little. The industry sells a retainer because the overhead of carrying a small client costs more than the work does, and the work itself genuinely does not recur.
All of it is now free
The method an operator used to pay for is published in full. An owner with an afternoon can do the whole thing themselves.
Twelve lessons, about seventy minutes, each ending with one question about your own property.
The visibility checkSeven questions, then a score and your gaps in the order worth fixing them.
City reportsWhat guests are asking AI about a given market, and what a page has to say back to be the answer.
The libraryThe reference material behind the lessons.
There is also a printable workbook: eleven sheets that, once filled in, say everything a machine needs in order to name a property.
