The cleaner sends two photos at 11:20. A wine glass in two pieces next to the sink, and a purple ring the size of a beer mat on the arm of the cream sofa. The next guest arrives at 15:00. You have about ninety seconds of clear thinking before the day takes over, and what you decide in them is worth more than either the glass or the stain.
Most hosts treat a damage claim as a question of principle: they broke it, they pay for it. It is not a question of principle. It is a trade, and both sides of it are real. On one side is the money you can recover. On the other is a review that sits on the listing for years, a guest who is about to be asked what they thought of you, and the fifteen to forty minutes it takes you to build a claim properly.
This article covers the arithmetic of that trade, the threshold below which you absorb the damage and say nothing, the evidence that makes a claim above the threshold survivable, the conversation that has to come before any platform process, and the repeated small damage that is not a guest problem at all.
The Arithmetic Most Hosts Skip
Put numbers on both sides. The ones below are illustrative: use your own, because the whole point is that your rate and your replacement costs are not mine.
The glass costs around four euros and came from a set you can still buy. The sofa arm is somewhere between sixty euros for an upholstery cleaner who gets it out and four hundred for a cover you cannot match. Those are knowable numbers. You can look them up in twenty minutes.
The other side of the trade is not knowable, and that is the part hosts get wrong. Nobody can tell you what a five-star review is worth on your listing. Anyone who gives you a figure is guessing, because it depends on your market, your season, your ranking and how many reviews you already have. What you can say with confidence is the shape: recovery is capped at the cost of the item, and a review is not capped in either direction. It affects placement and conversion for as long as it is visible, and a guest who feels accused writes a longer review than a guest who had a nice time.
So the four-euro glass is a trade you should never make. Not because four euros does not matter, but because the only ways it ends are "you absorb it anyway" or "you recovered four euros and put a review at risk." The sofa arm is a different conversation, and it deserves one.
There is a third cost that never appears in the decision: your time. Building a claim that holds up means finding the before-photo, writing the message, waiting, following up, and filling in a form. Forty minutes of your own labour has a price even if you never invoice it, and for small amounts it is often larger than the amount in dispute.
Set the Threshold Before You Need It
The decision gets made badly because it gets made at 11:20 with a cleaner waiting and a check-in in four hours. Make it once, in writing, per property, when nothing is broken.
A workable rule of thumb, again illustrative: absorb anything that costs less than one night at that property's average rate. A property at €120 a night absorbs the glass, the chipped mug, the missing tea towel, the bathmat, the lost remote. Above that line you look properly. Set your own number: the point is that the number exists before the photo arrives.
Then apply three tests to anything that clears the threshold.
Does it affect the next guest? A stained sofa arm four hours before check-in is an operational emergency regardless of who pays. Fix or mask it first, decide about money second. These are separate problems and mixing them is how the sofa is still stained at 15:00.
Is it damage or is it wear? A pan that has lost its coating over two years of guests is a cost of doing business. A pan with a burnt-on ring from one stay is closer to damage, and even then it may be neither worth claiming nor worth mentioning. Guests use things. Some of that use is what the nightly rate is for.
Would you raise it with a friend who had stayed? Not as a moral test, as a calibration. If the answer is no, absorbing it silently is almost always the better trade.
Evidence That Existed Before the Stay
If you decide something is worth pursuing, everything now depends on evidence you either have or do not. The most valuable piece is the one created before the guest arrived.
Keep a dated condition record per property: photographs of each room, close-ups of anything expensive or easy to mark, serial numbers for electronics, and a short written note of pre-existing marks. Refresh it at every deep clean, not once at listing. A record from eighteen months ago proves nothing about the state of the sofa in June.
The note about pre-existing marks does more work than hosts expect. It stops you claiming for something that was already there, which is the fastest way to lose both the money and the relationship, and it makes the rest of the record credible because it shows you are tracking the property rather than assembling a case.
Evidence Taken at Checkout
The second half is the checkout pass, and its value depends almost entirely on timing. Photograph before anything is moved, washed or straightened. A photo of a stain after the cleaner has scrubbed at it for ten minutes is a photo of a scrubbed stain.
Give the crew a fixed routine: wide shot of the room, then the close-up, then one shot that shows the damage in context so nobody has to take your word for which sofa it is. Same angles every turnover, so the before and after compare cleanly. The practice of building this into the standard pass is covered in how to use turnover inspection photos.
Two honest caveats. A photo pass adds real minutes to a turnover, and those minutes come out of a window that is usually already tight: know what your turnovers actually take before you add work to them, which is the subject of building your own turnover time baseline. And if anyone has entered the property between checkout and the photograph (a cleaner, a handyman, a next guest let in early), the chain is broken and you know it, so do not build a confident claim on it.
The evidence you gather is also a small pile of personal data about a specific guest, kept in a thread and a folder. Hold it while the dispute is live and delete it when the dispute closes; the retention logic is in what GDPR means for guest messaging.
The Conversation Before the Claim
The same claim lands completely differently depending on how it opens, and the opening is the part you fully control. A platform process is the last resort, not the first move: partly because it is slow, and mostly because a guest who first hears about the damage from a claim notification has been told that you did not think they were worth asking.
Get the timing right. Same day, not 07:00, and not while the cleaner is still sending photos. Wait until you have looked at the evidence and know what you are asking for. Three days later is worse: it reads as someone who went looking for something.
Open with a question, not a verdict. "The cleaner found a stain on the sofa arm this morning. Did something get spilled? No problem either way, I just need to know whether it needs professional cleaning before the next guest." That message does three things: it states the fact, it gives them a way to say yes without being accused, and it frames the cost as an operational necessity rather than a punishment.
Give them an exit. "If it was already there, tell me and I'll close it" costs you nothing and occasionally saves you from claiming for your own furniture. Most damage in short-term rentals is an accident that the guest already feels bad about, and a guest who is offered a dignified way to pay usually pays.
| How the message opens | What the guest hears |
|---|---|
| "You stained the sofa and left a broken glass." | I am being accused and charged. |
| "The cleaner found a stain this morning. Did something get spilled?" | I am being asked, and I can explain. |
| "Attached is an invoice for €180." | This is already decided without me. |
| "An upholstery clean is €90; I can send the quote." | This is a real cost, not a number he made up. |
| "This will affect your review." | This is a threat. |
The bottom row deserves its own sentence. Never mention reviews in a damage conversation, in either direction. It reads as coercion even when it is not meant that way, it is the sort of thing platforms take a dim view of, and it converts a recoverable situation into a public one.
Send one message, not five. Then wait. A guest who is drafting a payment does not need a follow-up two hours later.
When the Platform Process Is the Right Move
If the conversation fails, or the amount is large enough that goodwill is not the deciding factor, the platform process exists and you should use it rather than absorbing something serious.
Describe the shape to yourself before you start, and verify the current version rather than trusting any description of it, including this one. Broadly, the major platforms have generally operated damage or host-protection processes that expect you to have raised it with the guest first, that require evidence including proof of cost, that run on a deadline tied to checkout and often to the next check-in, and that involve the platform's own decision-makers if the guest declines. Every one of those elements has changed at some point, and the details differ between Airbnb, Vrbo and Booking.com, so check the current terms in your own account before you rely on a deadline you half-remember.
Two operational consequences follow from the shape. The deadline tied to the next check-in is why the checkout photo pass matters so much: a back-to-back turnover can close your window before you have looked at the photos. And filing generally notifies the guest, which ends the informal phase, so decide deliberately rather than filing as a reflex and negotiating afterwards.
Repeated Small Damage Is a Furnishing Decision
Here is the pattern worth more than any individual claim. If the same thing breaks at the same property three times in a season, it is not a guest problem. It is a furnishing problem wearing a guest costume.
Thin-stemmed wine glasses break. Glass-topped tables chip. Cream upholstery stains. Two hundred euros of matched crockery gets orphaned by one dropped plate. The claim recovers the item once; changing the item stops the loss and stops the conversation, and the conversation was the expensive part.
Keep a simple tally per property of what gets damaged and how often. Once a year, read it. Anything appearing three times either gets replaced with something that survives guests, or gets accepted as a consumable and priced into the nightly rate. Stemless glasses, wood instead of glass, a washable throw over the sofa arm, a second remote in the drawer. This is unglamorous and it beats winning claims.
The same log tells you something about the property rather than the guests: if the hall table is knocked over in every second stay, it is in the wrong place.
Honest Limits
Sometimes you lose both. A guest denies it, the platform declines the claim, and the review mentions that you tried to charge them: that outcome exists, it is not rare, and no process in this article prevents it. The reason to run the process anyway is that it improves your odds, not that it guarantees anything.
Platform resolution outcomes are not predictable. The same evidence pack can succeed one month and fail the next, decisions are made by people applying policies you cannot read in full, and hosts who tell you they know how to win these are describing a sample of one. Budget for the claim failing and you will make calmer decisions.
A guest who feels accused will say so publicly whether or not they were at fault. The tone advice above reduces how often that happens. It does not give you control over it, and a guest who was going to leave four stars over the parking will now leave four stars and a paragraph about the sofa.
There is also a limit on what any of this earns. The best-run damage process in the world recovers the cost of an item you have already lost the use of, and pays nothing for your time. Prevention (better furnishings, clearer house rules, a guest who feels looked after) pays better than recovery. That is also why a stay where problems were caught early tends to end without a dispute at all, which is the case for asking guests for feedback before checkout rather than discovering everything at 11:20.
And to be plain about the software: Welco does not file claims, does not arbitrate them, does not hold an evidence pack and will not tell you whether to pursue one. It is a messaging layer. Any vendor implying that automation resolves damage disputes is selling you the part that was never the hard part.
Where Welco Sits
Welco is a WhatsApp AI assistant for vacation rental hosts. Its part in this is upstream: verified guests can report a spill or a breakage the moment it happens, in their own language, and the report reaches you as a flagged escalation rather than a surprise at checkout, which is usually the difference between a conversation and a claim. How that fits the wider picture is in guest experience automation, and the review-side argument is in what AI guest messaging does to review scores. If you want to see how the escalation actually arrives, request a demo.