Four senior living operators, twenty-four communities, and one uncomfortable number: families choosing where a parent will live are doing it on fewer reviews than they would read before booking a restaurant.
A daughter has three weeks to decide where her father will live. She opens her phone and types four words.
Assisted living near me. What happens next commits a family to somewhere between fifty and a hundred thousand dollars a year, and commits a parent to the place he will most likely die. It is among the highest-stakes consumer decisions anyone makes.
And she will make the shortlist from three results on a map — a name, a rating, a distance, a review count. Before any tour, any brochure, any conversation.
4 operators are in this study. They are anonymised, and the letters are deliberately not in rank order — so if you work in this industry, there is a reasonable chance one of these scorecards is yours and you will not know which until you reach the end.
Read this as a list of things you can fix. Occupancy in this industry is won at the shortlist stage, and the shortlist is built on a screen before anyone picks up a phone.
Read this as a list of things to ask about. One operator here has its corporate office absorbing complaints meant for the company, and another is publishing a community under a brand it no longer uses.
Amplispot sells digital presence management, including a review and listings product called ReviewSpot. We have an obvious commercial interest in a report concluding that local presence is mismanaged, and you should read it with that in mind.
What we have done to make it checkable anyway: the scoring formula and full deduction schedule are published, so you can recalculate any score yourself. Every figure came from public Google listings, so any claim can be independently verified. No operator was contacted before publication, paid for inclusion, or given advance sight of its score. ReviewSpot is not mentioned again in this document.
Senior living is sold on the tour. Everyone in the industry knows this. What the industry has been slower to accept is that the tour is not where the decision narrows — the search result is.
A daughter comparing options forms a shortlist in under a minute from four things: whether the name reads as a real, specific place; whether the rating is above the two communities beside it; whether enough other families have been there for that rating to mean anything; and whether it looks open and active.
Each can be silently wrong. A community can trade under a name it abandoned years ago, so families searching the group brand never find it. A corporate office can be listed as a public destination and quietly collect complaints intended for the company. Two communities in one town can publish names so similar that reviewers confuse them.
Because of what the reviews contain. A restaurant listing carries opinions about dinner. A senior living listing carries families describing a parent's final months — sometimes gratefully, sometimes not. Those accounts sit permanently at the top of the result the next family reads, and they are written at the most emotionally charged moment of that family's life.
We have deliberately kept the substance of those reviews out of this report. What we are measuring is the infrastructure they sit on, not the care they describe.
None of this is visible from inside the building. Nobody on your team searches for their own community. That is the entire reason this report contains anything you did not already know.
If the shortlist is decided this early, the obvious question is how much evidence a family actually has to go on. The answer is startling.
Next, Part TwoThirty-eight reviewsAcross twenty-four communities we found an average of 38 reviews each. The thinnest operator in this study averages 27. Two individual communities carry fewer than ten.
Across the whole study. For comparison, multi-site veterinary clinics we audited on the same method averaged 154 to 1,628.
At that volume, three unhappy families visibly move a community's public rating.
Whose ratings are therefore statistically meaningless, while appearing on screen exactly as authoritative as any other.
Every operator here rates between 4.28 and 4.71. On sentiment, families are broadly satisfied and nobody is losing. But a 4.7 built on seven reviews is not a competitive advantage — it is a coin toss that happened to land well, and it will move the first time a family has a bad month.
Thin estates are fragile in both directions. They also rank poorly, because review count feeds local placement independently of score. An operator can be beaten in search by a lower-rated competitor with more families on record.
Four operators are the subject of this study. Before the numbers, here is precisely how we selected them and what our method genuinely cannot see.
Next, Part ThreeHow we lookedNo operator gave us data. None was contacted before publication. Every finding was equally available to the operator itself.
We searched each operator by brand name across the state or metropolitan area where it is most concentrated, so that communities are compared against genuine local peers.
For every communitie we captured the listing name, street address, telephone number, published opening hours, star rating, total review count, and the review text Google displays.
Most defects only appear across a whole estate — a name that differs from every sibling, a shared contact route, a listing with no reviews at all. This is why internal teams miss them.
Components were calculated from the captured figures using the formulas and deduction schedule in Part Five. No judgement was applied afterwards.
One operator in this study is doing something structurally different from the other three, and it is not what we expected.
Next, Part FourOne operator is not where you would expectThe operator that finishes first on our index does not have the highest member rating in the study. It wins on volume — 62 reviews per community against a study average of 38 — and on having no structural defects at all.
That is the whole argument of this report in one line. Sentiment is not the variable. Every operator here is well-rated. The separation comes from how much evidence exists and whether the estate is correctly represented.
The two lowest-scoring operators both carry a structural defect that has nothing to do with the quality of their communities. One publishes a community under a brand name it no longer uses, so families searching the group name cannot find it. The other has its corporate office listed as a public destination, where it has accumulated the lowest rating in its estate by absorbing complaints intended for the company rather than any community.
Both are administrative problems. Both are visible to every prospective family. Neither, as far as we can tell, is on anyone's dashboard.
On one operator's estate, identical review text appears verbatim on more than one listing, including the corporate office. We cannot determine from outside whether this is a disgruntled individual, a coordinated action, or something else entirely — and we are not going to speculate. What we can say is that it moves several ratings simultaneously and nothing suggests anyone is monitoring for it. If that operator contacts us we will send the specific listings immediately and free of charge.
That is one component of four. Here is the full measure, published so you can recalculate your own score.
Next, Part FiveA measuring stickAn index you cannot audit is a marketing device, not a measurement.
Reviews per communitie. Count is a ranking input in its own right, not just a confidence signal for humans. Log-scaled.
Whether the estate is correctly represented. Starts at full marks and loses points for each defect found.
Volume-weighted average across the estate, scaled 3.0 to 4.9.
The gap between the best and worst communitie. A wide spread usually means nobody is watching at communitie level.
Integrity begins at 35 points. Each defect costs the following:
Applied to 24 communities, that measure produces a table that looks very little like a ranking by size.
Next, Part SixThe ranking| # | Operator | Communities | Reviews | Per communitie | Rating | Defects | Score |
|---|---|---|---|---|---|---|---|
| 1 | Operator D | 6 | 369 | 62 | 4.67 | 1 | 81 |
| 2 | Operator A | 6 | 170 | 28 | 4.71 | 2 | 64 |
| 3 | Operator B | 5 | 192 | 38 | 4.32 | 2 | 58 |
| 4 | Operator C | 7 | 186 | 27 | 4.28 | 2 | 56 |
The largest operator in this study by community count finishes last. The operator in first place has no structural defects and the strongest review volume, and is not the best-rated. Scale gives you more communities; it does not give you a findable one.
Two variables drive nearly all of that separation. Neither has much to do with quality of care.
Next, Part SevenWhat the ranking revealsFrom 27 reviews per communitie to 62.
A more than twofold gap between the strongest and weakest estate. Because volume feeds local ranking independently of score, an operator at the bottom of this chart can be outranked by a lower-rated competitor with more families on record — losing the shortlist described in Part One before a tour is ever booked.
Take a location sitting at 4.6 stars. In one difficult month it receives three one-star reviews — a staffing gap, a bad week, one genuinely poor experience shared by a family or a group. Nothing unusual.
What that costs depends entirely on how many reviews were already on file.
| Reviews already on file | Rating after | Drop | Five-star reviews to recover |
|---|---|---|---|
| 40 | 4.35 | −0.25 | 27 |
| 100 | 4.50 | −0.10 | 27 |
| 250 | 4.56 | −0.04 | 27 |
| 500 | 4.58 | −0.02 | 27 |
| 1,000 | 4.59 | −0.01 | 27 |
Two things in that table matter more than anything else in this report.
The same three reviews do twenty-five times more damage at a thin location than a rich one. And the recovery cost — 27 five-star reviews — is identical at every volume. It does not get cheaper because you are large. It is fixed by arithmetic.
So a thin estate takes visible hits and pays the same repair bill. That is why volume behaves like insurance rather than marketing: you cannot prevent bad reviews, you can only dilute them, and dilution has to be running before the bad month, not after it. The average community in this study carries 38 reviews. That is the top row of this table — the most exposed position in the whole category.
From 4.28 to 4.71.
A 0.43-star spread across four operators. To a family comparing two communities that difference is invisible. This is what a category looks like where everybody has already solved the obvious problem, and where the remaining advantage lies somewhere else entirely.
Below is every operator, with its component scores and each defect we found, described generically.
Next, Part EightThe scorecardsWhat you take from this depends entirely on which side of the table you sit on.
Next, Part NineWhat to do about itIf any of the 4 scorecards looked uncomfortably familiar, there is a straightforward way to find out.
Next, Part TenWhich operator are youOperators in this study are anonymised. We hold a private key mapping each letter to its operator, along with a verification code that appears nowhere in this document.
Email us from a company domain and we will confirm your letter, quote your verification code back to you, and send the full underlying detail for your estate — every communitie, every defect, named and specific. No charge, no meeting required, and we will not add you to anything.
We will not confirm any other operator's identity to you, and we will not confirm yours to anyone else. If you believe a defect we recorded is wrong, tell us — corrections are published in the next edition with the correction noted.