This describes work that is in design. Everything below is what we are building, not something you can switch on today.
Thursday, 2:30pm, at a house nineteen minutes each way. Nobody comes.
The applicant is not rude. They booked four viewings on Saturday morning while scrolling, took the first unit they liked on Tuesday, and never thought about the other three again. That is not bad behaviour; that is what booking four things does.
So the slot is gone and the drive is gone — thirty-eight minutes of car plus the half hour you held. And here is the part that stings: the applicant who would have taken that house is sitting in your inbox right now, waiting on a reply to an email they sent on Wednesday.
Two halves of the same afternoon. One is empty and one is waiting. Nothing connects them except somebody noticing, and nobody noticed, because it was Thursday and there were other things.
The quietest loss in the business
Vacancy gets attention because it shows up in a report. This does not show up anywhere.
An empty slot leaves no artefact. Nothing is flagged, no number moves, and by Friday the Thursday hole has been overwritten by the week. If you asked the firm how much capacity it lost to no-shows last month, nobody could tell you — not because they are disorganised but because the question has no source.
It is capacity you had already paid for. The slot was reserved, the travel was committed, and the alternative use of that half hour was foreclosed the moment it was booked. A no-show does not cost you a potential half hour. It costs you a half hour you had already spent.
What we are building for it
Noticing. A showing was scheduled and did not happen. That is a fact with a time on it, and the moment it is known, something can act on it.
Recovering the applicant. They get a message the same afternoon, while the unit is still in their head — not a scolding, just the unit, and a link to pick another time if they still want one. People miss appointments for ordinary reasons and a fair number of them rebook when it is made easy. The ones who have moved on simply do not reply, which is also useful information.
Refilling the slot. The half hour goes back into availability, and the people most likely to want it are the ones who already asked about that property — which is what a per-property waitlist is for.
None of that requires anybody to remember anything on Thursday afternoon.
Why this is the best work on the roadmap
I want to make an argument here that runs against how software like this usually gets sold.
Look at what this actually needs. The trigger is already deterministic — a showing happened or it did not, and there is no judgement in that. The action already exists — the applicant re-books themselves through the same self-serve booking they used the first time. The whole job in between is one well-written message, with the right unit and the right link, at the right moment.
That is not a hard problem. It is barely an AI problem. And it is the best value-to-effort work in this entire programme, because the value is not in the sophistication of the response — it is in the fact that anybody noticed at all.
Ambition in this kind of software is not always a large model doing something impressive. Sometimes it is noticing that a firm's calendar has a hole in it on a Thursday afternoon and that two hundred metres away, in the same inbox, is the person who would have filled it.
The unglamorous dependency
Which brings us to the honest catch, and it belongs here rather than in a footnote.
None of this works unless somebody marks the showing as missed.
The software cannot infer a no-show. There is no signal — an applicant who does not attend does nothing, and doing nothing is indistinguishable from a hundred other kinds of nothing. Somebody who was there has to record that nobody came.
So the groundwork is attendance: a way of marking, in about two seconds, whether a booked showing happened. That is not an interesting feature. It is the cheapest possible piece of software and it is what makes everything above it nearly free — and it is the reason recording attendance sits ahead of the recovery work in the plan rather than being assumed.
It has a second payoff, which is that a firm that records attendance for three months can finally answer the question at the top of this piece: how much capacity did we lose, at which properties, in which slots. That is not a vendor's number about the industry. It is yours, about your buildings, and it is the only kind worth acting on.
You will not get a percentage from us. Not a no-show rate, not a recovery rate, not a reduction figure. This category is full of confident numbers on vendor pages that trace back to nothing, and we are not adding a fourth. What we can describe is the mechanism. What it does at your firm is measurable at your firm.
The other half: "any chance Thursday works instead?"
The same afternoon has a second failure in it, and it arrives in prose at eight in the evening.
"Sorry — any chance Thursday works instead?"
What happens now is four emails over two days. You propose two times. They can do the second but only after four. You check whether after four works. It does, mostly, and by the time it is settled neither party is completely certain what was agreed, and one of you writes "just to confirm" — which is the sentence that means the system failed.
What we are designing instead: the applicant writes in plain language, and gets real options back, from live availability, then and there. If it takes two or three exchanges, it takes two or three exchanges — that is normal and the design should expect it rather than collapse after one.
Two things about how that works are deliberate. The options offered are whatever the availability engine returned — the software presents them, it does not decide which applicant deserves which slot, and it never picks between people competing for a scarce time. And a reschedule that keeps going in circles stops being a scheduling problem: after a couple of attempts it goes to a person, because at that point something is going on that a calendar cannot fix.
A setting, not a philosophy
One trade-off worth naming.
Rebooking a no-show costs your firm nothing. It costs the slot something, and it gives an applicant a second chance they may not have earned. Some operators are relaxed about that — a missed viewing is a missed viewing, life happens. Others, particularly at scattered properties where every showing carries a drive, want a limit: one rebook, then you talk to a person.
Both are reasonable and they are not the same firm. So the cap belongs in your configuration rather than in our opinion. What matters is that it is a number you chose, not one you inherited from software.
Releaser is available to property management firms in the United States and Canada: one shareable link per property, your screening completed before anyone reaches your calendar, and self-serve booking for the people who meet the criteria you set.
No-show recovery and reschedule handling are in design. This is what we are building them to be.