What a One-Star Review Costs, and What Actually Repairs It

A single bad rating is not a score event. It is a booking event, and you can put an order of magnitude on it.

Last updated: September 20, 2026

A bad rating costs you bookings, not points, and the cost can be estimated in two steps: how far one review moves your average, which depends on how many reviews you already have, and how many straight top ratings move it back, which does not depend on your base at all. Between them sits the only decision that matters — which repair actions change the number and which only change how you feel about it. This guide works through the arithmetic, the repair sequence, and the point at which repairing stops being the right answer. Every figure below is an illustrative calculation, not a rule: platforms publish neither ranking weights nor rating thresholds, and both change, so read the numbers that apply to your own listing.

Key Takeaways

• The damage is dilution, and dilution depends on your base. Twenty reviews means one bad rating costs about a third of a point; a hundred means almost nothing.
• Recovery does not depend on your base. Nineteen straight top ratings undo one single-star review against a 4.8 target.
• A rating is a threshold worth clearing, not a score worth maximising. The value is in staying above the line, not in the second decimal.
• Two things move the average: new reviews, and time. Appeals, replies and discounts move other things, which is not the same.
• Stop when the complaint is structural. If three reviews name something you cannot change, the answer is repositioning, not repair.

What the rating actually gates

A rating is not a measure of quality that guests read carefully. It is a filter and a first impression, and both behave like a threshold rather than a slope.

What it touches	How it behaves	What you can control
Guest filters and sorting	A line to clear, not a gradient to climb	Whether you are above the line at all
Programme and badge thresholds	Set by the platform, shown in your own dashboard, subject to change	Knowing the current figure for your listing and not a second-hand one
The listing page impression	The average is read next to the price, in about two seconds	The three most recent reviews, which is what most readers actually see
Search position	Widely discussed by hosts; no platform publishes a weight	Nothing directly. Treat any specific weight you read as unverified

The practical consequence is that the number matters in two regimes and barely at all in between. Above the line, 4.8 and 4.9 are close, because the guests filtering you out have already stopped. Below it, the same gap is expensive, because you have dropped out of a set of results entirely. So the question after a bad review is not "how do I get back to 4.9" but "am I above the line, and if not, how fast can I get back over it".

Putting a rough number on one point of average

Two separate sums, and the second one surprises most hosts because it is the same number whatever size you are.

How far does it move? Your average is the total of all ratings divided by the count, so one new rating shifts it by less the more ratings you have. Starting from 4.8, with one single-star review added:

Existing reviews	New average	How much it costs you
10	4.45	About 0.35 of a point
25	4.65	About 0.15 of a point
50	4.73	About 0.07 of a point
100	4.76	About 0.04 of a point

A young listing is fragile in a way an established one is not, which is the best argument for asking satisfied guests long before you need one.

How much does it take to undo? Here the size of your base drops out. Each new top rating adds a small surplus above your target, and the bad review sits below it by a fixed deficit. To return to a target of 4.8 with all new reviews at five stars: the deficit is 3.8 points, each new review contributes 0.2 of surplus, and 3.8 divided by 0.2 is nineteen. One single-star review costs nineteen consecutive five-star stays to undo. Not nineteen in total — nineteen that land, one after another, with nothing else going wrong in between.

And what does that cost in money? Here the estimate gets honest about its own weakness: you need three numbers from your own operation and one assumption you cannot measure.

Line	Illustrative value	Where it comes from
Listing views per month	1,200	Your own dashboard
Booking rate	2%	Your own bookings divided by views
Monthly bookings	24	The two lines above
Conversion sensitivity to a rating drop	Assumed 5%–15%	An assumption. Platforms publish nothing, so bound it rather than find it
Bookings lost per month	1.2 to 3.6	24 multiplied by the assumed range
Contribution per booking	$150	Revenue minus the variable cost of the stay
Monthly cost while the rating is low	$180 to $540	Lost bookings multiplied by contribution
Recovery window at four reviews a month	About five months	Nineteen reviews divided by four
Illustrative total	$900 to $2,700	The monthly cost across the window

Read that last line as an order of magnitude, not a forecast. The sensitivity assumption is the weak link and nobody can supply it for you, which is why the output is a range. If the range is smaller than the repair you were weighing, do the arithmetic again before spending.

Those first three lines are the ones a dashboard should answer without a spreadsheet. You can put them in one place at localsbnb.com and connect a single channel before you commit to anything.

Repair sequence: what moves the number, and what only feels like it should

Exactly two things move an average: new ratings arriving, and enough of them arriving. Everything else is worth doing for other reasons, and confusing the two is how hosts spend a week on something that changes nothing.

Action	Moves the average?	Why
Fix the defect the review describes	Yes, but slowly	It stops the next nineteen guests writing the same sentence
Raise your review rate	Yes, and fastest	The recovery is nineteen reviews; the lever is how quickly they arrive
Reply to the review publicly	No	It moves the next guest, not the number
Appeal for removal	Rarely	Removals turn on policy, not on whether the review was fair
Offer the guest something to change it	No, and don't	Incentives tied to ratings are treated badly by platforms and read as pressure
Cut the price to win bookings back	No	Price moves demand; it does not move the rating, and discounting into a low rating spends margin twice
Delist and start again	No	You lose the base that was protecting you, which is what made the first one expensive

On appeals: outcomes are mixed, and decisions turn on each platform's content policy rather than on whether the review was accurate. One host's account in an Airbnb community thread dated 19 September 2026 described a guest who never stayed being allowed to leave a one-star rating, and a removal request refused on policy grounds. Treat that as one host's claim, not a rule. The lesson survives either way: an appeal is a request somebody else decides, while a new review is one you can go and earn.

The sequence that works is short. Fix what the review was about, so the next nineteen do not repeat it. Then raise the rate at which reviews arrive. Then wait.

When to stop repairing and start repositioning

Sometimes the review is not describing a defect. It is describing the property, correctly, to a guest who should never have booked it.

Signal	Repair or reposition	Why
Three reviews mention the same thing, and it is fixable	Repair	The complaint is specific and the fix is bounded
Three reviews mention the same thing, and it is the building	Reposition	Noise, stairs, no parking and a fourth floor are not defects
The fix costs more than a season of margin	Reposition	The arithmetic above is the ceiling on what you should spend
The guests who love it are not the guests who book it	Reposition	The listing is attracting the wrong reader
One complaint, nothing else like it in twenty reviews	Neither	Check for dilution: with a small base, one outlier is loud

Repositioning is not spin. It is changing the promise so the trade-off is visible before booking: leading with the stairs rather than burying them, naming the street noise and who it suits, pricing into the band where the compromise is understood. A guest who knew about the stairs and booked anyway does not write about the stairs.

The test for which side you are on is simple. Write the complaint in one sentence, then write what it would cost to remove it forever. If the second sentence has a number smaller than the cost you estimated earlier, repair. If it has no number at all, because the thing is the property, reposition.

FAQ

How many reviews do I need before a bad one stops hurting?

Look at the dilution table rather than a rule of thumb: at twenty-five existing reviews a single-star costs about a seventh of a point, at a hundred it is noise. What protects you is not a high rating but a large base, which is an argument for asking steadily.

Should I drop my rate until the rating recovers?

No. A lower rate buys demand, not ratings, and discounting while your rating is low means paying for occupancy with margin at exactly the moment your conversion is weakest. Fix the review pipeline, hold the rate, and let the average come back on its own.

One bad review costs nineteen straight good ones to undo, and the only two levers are the thing it complained about and how fast reviews arrive. Work out your own dilution, decide whether you are above or below the line, and spend accordingly. The revenue, occupancy and rate figures behind that decision belong in one dashboard too: localsbnb.com.

---

Every figure in this article is an illustrative calculation, not a forecast or a rule. Platforms do not publish ranking weights, and rating thresholds and review policies differ by platform and change over time: read the current figures that apply to your own listing. The community-thread account cited above is one host's claim, not a rule. Results vary by market, season and property. LOCALSBNB provides software, not financial or legal advice.