An AI visibility score can show how often your business appears in a set of generated answers. But an appearance alone does not establish that AI exposure contributed to a qualified inquiry or a booked appointment. A critical survey of GEO measurement makes a similar distinction: a citation in an answer does not, by itself, establish a source’s contribution (survey of GEO visibility measurement).
For an appointment-based business, that difference matters. A score may be useful for seeing whether your name shows up in answers to relevant questions. It is not a substitute for tracking the outcomes you care about: inquiries that fit your services and appointments that get booked.
What an AI visibility score measures
GEO, or generative engine optimization, is the practice of trying to improve how a business appears in answers from generative AI tools. A visibility score may count source appearances, citations or brand mentions in those answers, according to the GEO measurement survey.
The survey’s central point is that the questions used to calculate a score help define what the score represents. The authors call the selected situations and questions an “answer market.” They also note that the weights given to different questions affect the result. In plain terms, a score built from questions about emergency dental care may tell a different story from one built around cosmetic dentistry (survey of GEO visibility measurement).
Small changes in question wording can change which sources are retrieved and what appears in an answer. The survey also notes that, when a language model is asked to score an answer, the scoring instructions can change the score even if the answer itself stays the same (survey of GEO visibility measurement). That makes the score a result of both the questions you choose and the method you use to count appearances.
The paper is a critical survey and methodological framework, not a report of new experiments. Its authors say that the framework’s general empirical validity remains to be assessed (survey of GEO visibility measurement). So treat a visibility score as a useful monitoring signal, not as a proven measure of new demand.
Build a score around real buyer questions
You can make your tracking more relevant by choosing a stable set of questions that reflect the services and locations you want to grow. For a med spa, example questions might ask about options for a particular concern in a particular area. For a dental office, they might ask about a service, location and appointment need. These are starting points to tailor to your business, not a universal list.
Keep the wording consistent from one check to the next. Record the tool used, the question, the date of the check and whether your business was mentioned or cited. The survey identifies question wording, execution conditions, weighting and scoring rules as choices that shape a GEO measurement (survey of GEO visibility measurement). Recording those choices gives you context for interpreting changes in your own score.
Then track outcomes separately. For inquiries, note whether the person was a fit for your services and how they first heard about you. For appointments, record whether the inquiry became a booking. If you cannot reliably identify AI search as the source, avoid assigning every increase in bookings to a change in AI visibility. The score and the booking record answer different questions.
What current research can and cannot show
A second early research paper proposes a framework called Generative Marketing Mix Modeling to estimate the effects of generative search visibility and sponsored placements. A causal estimate is an attempt to separate an effect caused by a marketing change from one that simply happened at the same time. The paper’s framework compares expected business responses under alternative marketing sequences (Generative Marketing Mix Modeling paper).
For GEO, the proposed framework combines repeated generated answers with question counts, use across generative systems and estimates of whether people notice a business name. The authors investigate its performance using simulated product recommendation answers in English and Japanese (Generative Marketing Mix Modeling paper). That is an early methodological study, not evidence that a local clinic, salon or studio gained bookings from AI mentions.
What to do
- Choose questions that match your priority services and locations, then keep them stable for comparison.
- Monitor mentions and citations, but keep that score separate from inquiries and booked appointments.
- Use intake questions, booking records or other consistent tracking to understand where inquiries came from.
- When you change your marketing, compare outcomes in a way that can help distinguish the effect of the change from other activity happening at the same time.
The practical goal is not a perfect AI score. It is a clearer view of whether visibility is connected to the inquiries and appointments your business needs. If you want help testing marketing lift, Vetraeus Value Attribution is designed for experiments that measure it.
- ai search
- marketing measurement
- appointment bookings
- local marketing
- value attribution
Alan Jordan, founder and principal solutions architect. Data engineering, data science and business analytics for appointment-based businesses.
