Consider a group like this one, a pattern that appears repeatedly across The Dental Index data. Daniel Mensah runs operations for a 14-location dental group spread across three states. New patient volume has been flat for two straight quarters, even though paid search spend is up almost a third. The chairs are full. The hygiene columns are booked out. Every location looks like it is working, but the growth curve has gone sideways and nobody in the weekly numbers can say why. The clinical operation is not the problem. The pipeline that never forms is. If your group is spending more to acquire patients while the new patient line refuses to move, you may be looking at the same pattern.
Patients do not open ChatGPT and type "find me a dentist." They describe their situation. "I have been putting off implants for a year and I am finally ready. What should I ask when I call?" "My gums bleed every time I brush. Is that serious enough to see someone?" The AI reads the dental content it can find, builds an answer, and sometimes, inside that answer, it names a practice. That is the moment that decides your new patient number, weeks before your front desk ever picks up.
There are 432,000 AI-assisted dental searches every month in the United States. Your portfolio sits inside that demand pool, or it does not. The difference is not your acquisition budget. It is your positioning, and right now it is expressing itself in the flat curve Daniel cannot explain. The patients are searching and deciding, just somewhere your locations are not named.
What Are Patients Actually Asking ChatGPT Before They Book?
The conversation that brings a patient to your front desk starts weeks earlier, in a private exchange with an AI assistant. The patient is not comparison shopping yet. They are processing anxiety and quietly looking for permission to make the call. That is a different state of mind than a directory click, and it produces a different kind of patient. The questions fall into three buckets, and each tells you what the patient is really doing.
Procedure validation questions: "Is a full arch implant worth it at my age?" "How long does Invisalign actually take for adults?" The patient has half-decided and wants a reason to believe the decision is sound. The practice the AI names becomes the default place to act.
Fear reduction questions: "Does getting a dental implant hurt?" "What happens if I have severe anxiety and need a crown?" By the time they call, the fear has been worked through in a conversation you were not part of. They arrive ready.
Qualifier questions: "What should I look for in a dentist who specialises in implants?" The patient is building the checklist they will judge you against. If your locations match the description the AI just gave them, you are not being evaluated. You are being confirmed.
When the AI answers, it pulls from content it can find, parse, and trust. If a location has written clearly about those exact situations, in the language patients use, tied to a specific service and a verifiable address, the AI includes it. If it cannot read that signal, the location is not in the answer. Your 14 locations are in those answers, or they are not, and the data says most are not.
Why Does the AI Name Some Practices and Not Yours?
Seventy percent of dental practices are invisible to AI search. Read that against a 14-location portfolio and the math is uncomfortable: roughly ten of your sites do not appear when a patient asks an AI assistant for guidance before booking. Your new patient numbers carry that weight right now, disguised as a flat curve and a rising acquisition cost.
The practices that do get named share one structural trait. They publish content that mirrors how patients describe their problems, not how clinicians describe procedures. A patient asks about "the back tooth that cracked on a popcorn kernel," not "a fractured second molar requiring a crown." The named practices speak the patient's language and tie it to a verifiable local identity: a complete Google Business Profile, consistent name and address signals across every directory, and reviews that mention the actual procedures. That coherence is what lets an AI recommend them with confidence. This is the entire substance of AI integration at scale, and it scales: what makes one location legible to AI is what makes fourteen legible, applied with discipline across every market.
Most groups misread this as a budget question and raise the paid search spend, which is what Daniel did. But the AI does not read your ad budget. It reads your signal. A location can outspend every competitor in its market and still be absent from the answer, because the answer is assembled from positioning, not promotion.
How Big Is the Gap Across Your Locations, in Real Numbers?
The average demand capture rate, the share of available patient demand a practice actually converts, sits at just 2.3% even among top-ranked practices. Fewer than 8% of US practices score above 65 out of 100 on AI readiness, and the national average sits below 40. Those numbers describe the spread between the locations in your group that grow and the ones that quietly stall, which the table below shows when you sort your portfolio by readiness rather than collections.
| Location AI Readiness Tier | Share of Practices Nationally | AI-Referred Engagement | High-Value Booking Pattern |
|---|---|---|---|
| High readiness (above 65/100) | Fewer than 8% | 7x more AI-referred clicks with a complete GBP | Books implants and cosmetic cases at 2-3x the rate of other channels |
| National average (below 40/100) | The majority of practices | Roughly 70% invisible to AI-referred patients | Flat new patient curve, capture stuck near the 2.3% average |
| Mixed portfolio (uneven by site) | Most multi-location groups | Engagement concentrated in one or two strong sites | A few locations carry the group; the rest absorb the spend |
Source: The Dental Index national practice audit · 2026
Look at the bottom two rows, because that is where most groups live. If your growth is carried by two or three locations while the rest run full but flat, you are reading the mixed-portfolio row in your own numbers. The strong sites are not strong because of better dentistry. Their positioning simply happens to be legible to AI. The flat sites do the same clinical work and get none of the credit, because the patients researching their next implant never see them named.
This slice matters because of what the patient has already done before calling. AI-referred patients book high-value procedures at two to three times the rate of standard-directory patients, because the AI conversation works as a pre-consultation. By the time they dial, the decision is narrowed to acting. They call to confirm, not to be sold. With implants averaging $4,500 and cosmetic cases $3,800, a location absent from AI answers is missing the exact slice that pays for everything else.
The patient who arrives through AI search is not comparing anymore. They are confirming.
What Do the Groups Getting This Right Do Differently?
They stop treating AI visibility as a campaign and treat it as infrastructure: built once and maintained at every location, the way you maintain sterilisation standards. The groups that win the AI answer have made positioning a managed function, with someone accountable for whether each site is legible to the systems patients now use to choose a dentist.
That reframe changes what they measure. Instead of asking only "how many calls did we convert," they ask "in how many of our markets does the AI name us when a patient describes the case we want?" That question has a measurable answer for every location, and it predicts the flat curve before it reaches collections. The difference between the strong sites and the stalled ones is whether positioning is controlled deliberately or stumbled into.
Consistency is the lever scale gives them. A solo practice has one Google Business Profile to perfect. You have fourteen, and a group that standardises legible positioning across every site compounds visibility the way a single practice never can. The same pattern that explains why high-visibility practices grow faster also explains why those practices carry lower overhead: the patient mix AI delivers changes the revenue side of every location's economics.
What Separates the Groups That Fix This From the Ones That Don't?
It is not budget, and it is not a new vendor. It is a change in how the operator sees the problem. The groups that stay stuck file AI invisibility under marketing, where it competes with ad creative and never gets owned. The groups that fix it move it to operations, where positioning becomes a standard every location is held to.
That single move changes everything downstream. Once AI legibility is an operational standard rather than a marketing wish, it gets a metric, an owner, and a place in the weekly review. The fix was never a bigger spend. It was a decision to treat the conversation before the call as part of the business.
Daniel's group had been doing the opposite without naming it. Every location optimised hard for the moment a patient called, and not one was accountable for whether a patient ever heard the practice named in the first place. They had built a sophisticated second act and no first act at all.
How Long Before You See a Difference?
The signal changes fast. The numbers follow on their own clock. Completing a location's Google Business Profile, building procedure-depth content in patient language, and aligning the review signal are changes you can make in weeks. What takes longer is the patient mix responding to them.
Practices improving their AI readiness typically begin to see a shift in case mix within two to three months, as the patients researching high-value procedures start finding them in the answer. The new patient curve usually registers a measurable change around the six-month mark and compounds from there. At portfolio scale the aggregate effect arrives in waves as each site clears the threshold. You are not waiting on one curve. You are waiting on fourteen, staggered by how legible each one already is.
What Changes When You Treat the First Act as Part of the Business?
Six months after Daniel's group reframed the problem, the pattern in the numbers had moved.
The flat curve was no longer flat.
The locations that had been stalling were appearing in AI answers for the procedures they were built to do.
Implant and cosmetic cases were taking a larger share of the schedule across the portfolio, not just at the two sites that had always carried it. The acquisition spend had not gone up again. In several markets it came down.
Nothing in the clinical operation had changed. The same dentists were doing the same work in the same chairs. What changed was whether the patients deciding on their next implant could find those chairs before the decision was made. If your group is running full and flat, the mechanism is the same, and so is the place it starts.
Where Your Positioning Either Reaches Patients or Disappears
Every argument here ends in the same place. Clear positioning makes your locations grow, but only if it is visible where patients are looking, and in 2026 they look inside an AI conversation first. A group with strong clinical positioning and weak AI visibility is, to the patient researching a high-value case, simply absent. For a multi-location group, AI search is not one more channel. It is the infrastructure through which your positioning surfaces in every local market at once. Capturing those monthly searches across fourteen markets is not a creative problem you solve with a better ad. It is an operations problem you solve by making each location legible, consistent, and named. The groups that treat it that way pull ahead of the ones still buying clicks on the channels they already own, and the gap widens every quarter because visibility compounds.
You have spent years building locations worth choosing. The only question is whether the patients who would choose them can find them in the conversation that now decides. Find out where each location stands in AI search and Google Maps, and you will see where your flat curve begins.
Treat the conversation before the call as part of your business
The patient who eventually books has already decided in a conversation you were not part of. The groups that win those patients stop designing only for the front desk and start designing for the AI exchange that shapes the decision first. Accept that the first act is already happening, and you stop losing patients you never knew you were in the running for.
Speak the patient's language at every location, not the clinician's
Patients describe a cracked tooth, a year of putting off implants, the fear of the drill. They do not describe fractured molars and osseointegration. The AI names the locations whose content mirrors how patients actually talk, because that is what it matches a patient's question against. A site written in clinical vocabulary is invisible to the patient who never uses those words.
Make positioning a managed standard across all your sites
A patient researching an implant trusts a group whose identity is consistent everywhere they look. When one location says one thing and another says nothing, the signal weakens and the AI hesitates. Standardising what each site is known for, market by market, turns scale into an advantage a solo practice can never match.
Sort your portfolio by visibility, not by collections
The patient never sees your P&L. They see which location the AI names when they describe their case. Ranking your sites by whether they appear in those answers tells you which locations are pre-selling high-value patients and which are running full but invisible, months before the gap reaches the management report.
Align the review signal with the cases you want to win
When patients read reviews that name implants, full-arch cases, and anxiety-free care, they arrive already believing a location does that work, and they accept treatment accordingly. When the reviews only mention friendly staff and easy parking, the AI has nothing to match a high-value question against. The review language a location collects decides which patients believe it is for them.
Know your AI readiness score before a competitor names the gap for you
A patient asking an AI assistant which group to trust is reading a public record of your visibility, whether you have looked at it or not. Knowing where each location stands tells you which markets are pre-selling your highest-value cases and which are handing them to the practice that bothered to be legible. See where your locations stand at javeriarnaqvi.com/dental-group-growth-consultant.