A cancelled 2pm cleaning has an expiry date, and it’s roughly lunchtime. After that, nobody on your waitlist can rearrange their afternoon. They’ve got kids to pick up, a shift to cover, a boss who already said no once this month. The slot is still sitting on the calendar, technically open, but it’s already dead.
Our demo already handles the obvious part. When someone cancels, it pushes the empty slot out to people who want an earlier appointment, and it works. But “it works” isn’t the same as “it works as well as it could,” and the gap between the two comes down to timing and who gets asked first. That’s what part two of our Clinic Scheduling & Recovery series is about.
Why does timing matter so much?
Every cancelled slot sits somewhere on a curve. With three days’ notice, it’s easy to fill, because plenty of people can plan around it. With four hours’ notice, you need someone who’s retired, works nights, or happens to be sitting at home with a toothache. Somewhere on that curve there’s a cliff, a point where the chance of filling the slot drops off sharply and it stops being worth the effort, or the text messages. We don’t know where that cliff is yet. Nobody at a small clinic really does, and that’s exactly why it’s worth measuring.
What are we trying to find out?
The question we’re asking is how the time between a cancellation and the backfill offer, and the way the waitlist is ordered, affect the chance that a cancelled slot gets rebooked. To answer it, we’ll measure fill rates across four notice windows: under four hours, four to twenty-four hours, one to three days, and more than three days. We’ll compare different ways of ordering the waitlist, look for the point where offering a slot is no longer worth it, and measure chair utilization before and after backfill is switched on, so we can see the effect on the clinic’s day as a whole.
Who should get the offer first?
This is a surprisingly human question dressed up as a technical one, and we’re comparing four answers to it. First come, first served is fair, simple and what patients expect. Closest by distance makes practical sense, because someone ten minutes away can say yes to a same-day slot while someone forty minutes away probably can’t. Highest value treatment is good for revenue, but it can feel a bit cold if a patient who’s waited three weeks keeps getting skipped. And longest waiting is the kindest option, though maybe not the fastest. We suspect the answer isn’t a single rule. It’s probably closest person first for same-day slots, and longest waiting for anything with more notice. But that’s a guess, and we’d rather have data.
How are we testing it?
The first step is to log every cancelled slot along with exactly how much notice was given. Then we record every offer that goes out: who got it, on which channel, in what order, and when they replied, if they replied at all. To compare ordering rules fairly, we’ll alternate them week by week at the same clinic, so the staff and patients stay the same and only the rule changes. Utilization is calculated as booked chair minutes divided by available chair minutes for each day. Finally, we’ll plot fill rate against notice window. We expect a sharp drop somewhere on that chart, and that drop is the finding.
What data does this need?
The demo needs the time each cancellation happened and the time of the original slot, which together give us the notice window. For every offer, it stores the recipient, the channel, the send time, the reply and how long the reply took, so we can see how fast people actually respond. It records whether the slot was filled and with what kind of appointment, and it tracks the total available chair minutes per provider per day so utilization can be calculated. It also captures the size of the waitlist at the moment of each cancellation. That detail matters more than it sounds, because a slot with two people waiting is a very different problem from one with twenty.
What would surprise us?
It would surprise us if waitlist patients grabbed same-day slots far more often than we expect. Retirees and shift workers often have flexible mornings, and if our data shows they’re the ones quietly saving the clinic’s afternoons, that’s not just a scheduling insight. It’s a marketing one. A clinic could actively build a “call me if anything opens up” list from exactly those people. In the first post, we looked at predicting cancellations before they happen. Next up is WhatsApp, email or SMS, and which channel patients actually reply to