Picture a family member calling an assisted living community about an elderly parent. By the time they pick up the phone, they’ve already asked ChatGPT which communities nearby handle memory care well, cross-checked the two names it gave them against Google reviews, and picked the one that came back clean on both. The community has no idea any of that happened. As far as its CRM is concerned, this is a first-touch inbound lead.
The real bottleneck is operational, not technical
A single-location practice can put one person in charge of one Google Business Profile and one set of reviews, a solvable, if tedious, job. A DSO or senior living operator built through acquisition inherits something harder: dozens or hundreds of locations, each drifting out of sync in its own way. This is fundamentally an operations problem, not a technology one.
A typical acquisition-built portfolio looks like a patchwork:
- Some locations with a profile one local manager has kept up for years
- Others still running under the acquired brand’s old name
- A few with no one clearly responsible for either
AI didn’t create that fragmentation. It just made the cost of it visible in a new place, since a chatbot summarizing the organization now has to make sense of the same inconsistency a human researcher eventually would have noticed anyway.
A website rebuild carries the same exposure: a typical migration sheds five to twenty percent of a site’s ranking keywords, which for a multi-location organization is this exact fragmentation risk showing up at the moment local visibility is most exposed.
One multi-location DSO client used a shared SEO playbook across five brand website launches in a single year. None of the mature launches lost their ranking footprint through migration. On the largest brand, more than 2,700 keywords gained ground after launch, compared with roughly 750 that declined.
These results demonstrate the value of a repeatable process for protecting and improving search visibility across brands. They do not establish whether the organization appeared more often in AI recommendations. That requires separate measurement.
The confirmation call
Call it the confirmation call. The conversation that used to start the relationship often confirms a shortlist formed elsewhere. It’s happening across every category of multi-location healthcare: dental service organizations, senior living and assisted living operators, physical therapy networks, home health agencies.
A 2026 Pew Research Center survey found 34% of U.S. adults have used an AI chatbot for at least one health-related task, most often to get quick information or figure out what might be causing a symptom.
Health questions carry a different trust bar than most searches. Nobody fact-checks a coffee shop recommendation the way they fact-check a recommendation for their mother’s care. That higher bar cuts both ways: people are more likely to cross-reference an AI answer against something else, and more likely to anchor hard on whichever source sounded most authoritative first. Organizations are now competing to be that first authoritative-sounding source.
Reviews still matter, but a machine reads them first
BrightLocal’s 2026 Local Consumer Review Survey, a panel of 1,002 U.S. adults, found that AI tools jumped from 6% to 45% of local business discovery in a single year.
An AI system doesn’t read reviews the way a person browsing a profile does. A person skims toward the top, notices the freshest reviews, and discounts anything that reads like old news. A model synthesizing sentiment across everything indexed has no equivalent instinct to let old complaints fade.
A billing dispute from eighteen months ago that a practice resolved can still shape how a chatbot describes that location today, long after a human reader would have scrolled past it.
Balancing AI visibility and location-level search results
For a single-location practice, that’s the whole battle: win the AI layer, win the review layer, done. Multi-location organizations fight on a second front at the same time, and it’s the one that actually decides who gets the call.
More than half of consumers, 52%, started their most recent local business search directly on Google Search itself, according to BrightLocal’s Consumer Search Behavior report. That’s the layer AI hasn’t touched.
Brand-level AI visibility, what a chatbot says about the organization as a whole, is a separate problem from location-level local search, what actually shows up in the map pack when someone searches “physical therapy” in a specific zip code.
What this looks like for dental groups
Dental consolidation makes the brand-versus-location problem sharper. The American Dental Association’s Health Policy Institute found that 27% of dentists less than 10 years out of school are now DSO-affiliated, up from 24% the year before, a share that keeps climbing as more practices join larger groups.
Patients built loyalty to a dentist over years of visits, not to a logo. When that practice joins a DSO, the corporate parent inherits the address and the equipment, not the trust. Most rebrand plans handle the legal name change and the new signage and stop there, treating trust as something that carries over automatically.
It has to be rebuilt in the exact content, the location page, the provider bio, the specific-sounding detail, this piece keeps coming back to. A page built from a template with the city name swapped in reads exactly as generic to an AI system as it does to the patient.
What this looks like for senior living and home health
Senior living operates on the same logic, with a twist: the research often starts before anyone is in crisis. A 2026 U.S. News & World Report survey found 65% of independent seniors start researching senior living proactively, specifically to keep their families from having to make an emergency decision later.
But not every search in this category is equal:
- A search that starts calmly looks like comparison shopping: cost, amenities, reputation, weighed over weeks
- A search that starts at 11 p.m. after a call from a hospital discharge planner looks like triage: find something that sounds safe, fast
Both queries hit the same AI tools and the same review platforms, but they’re reading for different signals, and most location content is written only for the first kind of visitor.
Home health care decisions are often made on short notice. CMS built the Care Compare tool specifically so families could compare home health agencies on public quality data, and U.S. News added its own home health ratings in 2026 for the same reason it gave for senior living: these decisions often happen during a health event, on short notice.
The window to act is also shrinking. NIC MAP data shows senior housing occupancy near record highs, with new construction at its slowest pace since NIC began tracking supply in 2006. Fewer openings means the family that gets an accurate answer first is the family that gets the room.
Don’t overcorrect toward the algorithm
The instinct here is to chase the AI answer and let the fundamentals slide. The data argues against that.
Direct Google Search still accounts for the majority of where local searches start. Winning the AI layer while ignoring basic Google Business Profile hygiene at each location leaves the organization sounding smart in an answer that nobody who was actually going to call ever saw.
Assign ownership, then fix the gaps
The first decision is who keeps location information accurate. The central marketing team should own the shared standards, cross-location audits, and reporting. Each location should have a named owner responsible for confirming providers, services, hours, and availability, with a clear process for getting changes published.
Leadership should review which locations have unresolved information gaps, who owns each fix, and how long those gaps remain open. Track AI visibility alongside local search performance, calls, and appointments; report each separately so visibility changes aren’t mistaken for patient growth.
Start with five checks: AI answers, review activity, Google Business Profiles, location pages, and availability information.
- Audit your AI footprint before you optimize anything else. Query ChatGPT, Gemini, and Google’s AI Overviews with the exact phrases patients and families actually type: “best [category] near [city],” “is [organization] any good,” “[organization] vs [competitor].” Run this for the brand name and for your five highest-volume locations, log where each tool is wrong, outdated, or silent, and rerun the same queries monthly, because these answers shift as models retrain and as your own content and reviews change. Treat this as a standing metric next to organic rank, not a one-time gut check.
- Give reviews a cadence, not a campaign. Set a minimum-velocity target per location, for example five new reviews a month, and a recency ceiling so no location goes more than 60 days without a fresh one. Assign ownership to one specific team, and build the ask into an existing operational moment, discharge, checkout, program completion, rather than a separate email blast, since reviews tied to a real visit read as more credible to people and AI systems alike than a batch solicited all at once.
- Run the Google Business Profile audit location by location, not brand-wide. Pull every location into one sheet and check five things for each: primary category accuracy, photo freshness, completeness of attributes like telehealth or accepts-new-patients, NAP consistency against the website and any lingering acquired-brand listings, and review response rate. Fix the worst-performing locations first, because a handful of neglected profiles can drag down how an AI system characterizes the entire organization, then put a recurring quarterly audit on the calendar so the fixes hold.
- Replace templated location pages with pages a real person could have written. Name the actual providers at each site, describe what genuinely differs there, specialties on staff, languages spoken, parking, accessibility, hours, and add at least one FAQ pulled from questions your own front desk or intake team actually fields locally. Cut the boilerplate paragraph that repeats across every location with only the city name changed, since that’s the easiest thing for a patient or an AI system to spot as generic. For our DSO client, rebuilt location pages on the most recently launched brand saw clicks rise 6.6% and impressions rise 23.5% within two months of launch, and the site overall saw appointments scheduled rise 49% and phone calls rise 43.5%, period over period. Real local detail pays off like that once an organization actually builds it.
- Keep availability signals accurate in real time wherever capacity is genuinely limited. For senior living, home health, and any other vertical with real openings to track, connect your CRM or scheduling system’s live availability to whatever feeds your Google Business Profile, website, and any AI-facing data source, so no family is told a room or slot is open that filled days ago. Where a real-time feed isn’t feasible yet, set a manual update cadence of at least once a week, and treat any location that goes stale past that window as a lead-loss risk, not a bookkeeping backlog.
What’s next
Multi-location healthcare organizations spent the last decade building marketing that could scale: centralized paid media, shared creative systems, one CRM instead of forty. The AI layer is the next thing that needs the same treatment.
Level is publishing a closer look at dental service organizations and senior living later this year. Worth watching for, especially if intake at your organization still can’t say what an AI told a patient before they ever called.
