Worked example — not a client result. Every practice, figure and timeline on this page is an example built from a typical engagement. It shows how the work runs and what it produces. It is not a claim about a named client, because there is no client behind it.
The Practice
An emergency and walk-in dental practice in Denver: two dentists on the chair, a door that is open when other dental offices are shut, and a reputation built almost entirely on word of mouth from patients who had a bad night.
The practice was busy. It was not invisible in the way most practices are invisible. It was invisible in one specific, expensive way: for the questions that happen at eleven at night.
The Challenge
Emergency dental questions are the highest-intent questions anyone asks a dentist. The patient is in pain, the clock is late, and the decision takes minutes rather than weeks.
That made the gap expensive. Across the eighteen high-intent emergency questions in the tracked set, the practice was named three times. Three practices up the road were named in the same answers, every time.
Nothing the practice owned was wrong. The site was accurate, the hours were right on the website, and the reviews were excellent. The problem was that an AI engine answering "emergency dentist in Denver at midnight" had nothing specific to quote. The practice had one page that mentioned emergency appointments inside a paragraph about general dentistry.
An engine cannot quote a paragraph. It can quote an answer.
What We Did
1. Published The Answer The Patient Actually Needs
One page, written for someone in pain at midnight: what counts as a dental emergency, what to do in the first hour, what it costs, what insurance covers, who to call, and where to go at two in the morning. Six pages underneath it, one per emergency - knocked-out tooth, severe toothache, swelling, broken tooth on a holiday, dental trauma, and what to do when the pain is in the jaw.
Each page answered its question in the first paragraph and then explained it properly. Headings were the questions. Nothing was buried.
2. Fixed The Facts The Engines Read, Not Only The Ones The Site Owned
The same practice name, address, phone and hours in every directory, and the hours corrected in the two listings that still showed a 5pm close. This was unglamorous work and it was the single biggest change in the measurement: the engines were quoting "open until 5pm" for a practice open until midnight.
3. Made The Entity Unmistakable
Dentist and MedicalBusiness markup with the real services, the real hours, the real coordinates and the real emergency availability. The markup matched what a patient sees on the page. No two versions of the truth.
4. Measured The Emergency Set Every Week
The same eighteen high-intent emergency prompts, the same four engines, the same city, every week. We recorded whether the practice was named, where it appeared, which competitors were named alongside it, which pages were cited, and whether the facts the engine gave were correct.
The last check is the one that pays: two engines continued to quote a competitor's Saturday hours for six weeks after the practice had corrected its listings. The fix was in the directory, not on the website.
Results After Eight Weeks
Swipe the table sideways to see every column.
| Metric | Before | After |
|---|---|---|
| AI mentions on emergency prompts | 3 | 17 |
| Calls from unbranded search | 62 | 94 |
| Emergency pages published | 1 | 7 |
| Competitors named on the same prompts | 6 | 4 |
These figures are counts across a fixed prompt set. They are not a ranking, not a recommendation, and not a promise of one.
The Takeaway
The practice did not need more patients to know about it. It needed to be the answer to the question its best patients were already asking.
An emergency question has a short half-life. If the practice is not in the answer, the patient is in someone else's waiting room, and no amount of reputation reaches them.
How This Runs For You
This is the monthly work in Dental AEO, and the free snapshot is the first cut of it: ten questions, four engines, about a minute - and you see exactly where you stand before you decide anything.