Should we double-book to cover no-shows?

The idea usually arrives from whoever runs the schedule, at the end of a month with too many empty chairs: if we already know roughly one in eight won't show up, why not put two people in some slots and let the averages sort it out? Airlines do it. Restaurants do it. It sounds like free money sitting on the table. It is worth doing the arithmetic before the policy meeting, because the numbers are not close.

What overbooking is actually betting on

An airline overbooking a 180-seat plane is pooling risk. Every no-show anywhere on that plane absorbs any extra passenger, because a seat is a seat. Your schedule pools nothing. The patient who no-shows at 2:40 does not absorb the extra person you double-booked into the 9:20 — different chair, different hour, different provider, often different room. So you are not betting on how many people miss, which you can estimate reasonably well. You are betting on which slot goes empty, which is close to a coin toss you have to call twice a day, every day, forever. That distinction is the whole article.

The arithmetic, with example numbers

Take a practice running 32 appointment slots a week with a 12% no-show rate — an example, use your own figures. That is about 4 empty slots a week, so the plan is to double-book 4 slots to cover them. Work out what happens in those 4 slots. Both patients show: 0.88 x 0.88 = 77% of the time, which is 3.1 collisions a week. Exactly one shows — the case you were hoping for, where the double-book quietly saves you: 2 x 0.88 x 0.12 = 21%, which is 0.8 recovered slots a week. Both miss: 1.4%, or one slot every seventeen weeks.

So the trade is roughly three collisions for every one slot recovered. At an example slot value of $180, the recovery is worth about $150 a week — and it costs you three weekly incidents of two patients in a waiting room built for one appointment. Meanwhile the other 28 slots still no-show at 12% and the four empty chairs you were trying to fix mostly still happen, because you guessed the wrong slots. Change the no-show rate and the ratio moves, but not in your favour: the lower your no-show rate, the more collisions you buy per recovered slot. Overbooking works best exactly where it is least needed.

What a collision actually costs

A collision does not appear anywhere on the schedule, which is why this idea survives so long. It shows up as a patient who waited forty minutes past their appointment time, a provider running behind for the rest of the afternoon, a front desk absorbing the complaint, and a visit that got compressed to catch up. The two patients most likely to be harmed are the two who did exactly what you asked — they showed up, on time, for a slot they were promised. And the practice pays for it twice, because a rushed, late visit is precisely the experience that produces the review you then have to answer.

The version of overbooking that is defensible

Blind double-booking spreads a bet across patients at random. Targeted standby does not. Two things are genuinely worth doing: name the slot, not the day. If a specific slot has a measured pattern — the 8:00am first-of-the-day, or a particular repeat offender — that is a data-backed bet on a specific chair, not a policy applied to everyone. Overbook the waiting list, not the chair. Keep a live short-list of patients who asked for anything sooner and text the slot the moment it is confirmed dead, rather than pre-committing a second body to a chair that might well be occupied. Same recovered revenue, zero collisions, and nobody waits forty minutes for a slot that was theirs.

Do the cheaper thing first

Every no-show you prevent returns a full slot with no collision risk at all — which makes prevention strictly better than overbooking at every no-show rate. The prevention work is unglamorous and mostly mechanical: instant confirmation at the moment of booking, a reminder ladder that carries a tappable reschedule path so a Wednesday cancellation becomes refillable inventory instead of a Thursday empty chair, and something that actually fills the gap when one opens. In a live business we measure, about 1 in 3 inquiries entered by staff lost their next step entirely — not because anyone was careless, but because the step lived in someone's memory rather than in the system. Booked appointments leak in the same place, and that leak is far cheaper to close than it is to hedge against.

Score the schedule before you gamble on it

Before you decide whether to double-book, it is worth knowing which step of your booking-to-chair flow is actually leaking. The free score takes three minutes: ten questions about your flow, scored 0–100, your weakest step named, and the monthly no-show cost estimated from your own numbers with every assumption labeled. If the score comes back with a missing confirmation step or no backfill path, you have found something worth more than any overbooking policy — and the $500 first fix baselines the flow and installs the missing step, timestamped.