Why appointments run late in a pattern you can predict

Delay is cumulative, and that is the whole story

A booked session is a sequence in which each appointment begins when the previous one ends. Nothing in that arrangement gives back time. If an appointment overruns, the next starts late; if it also overruns, the lateness compounds; and there is no mechanism anywhere in the sequence for the schedule to recover except an appointment finishing early or somebody failing to attend.

That asymmetry produces the pattern everybody has noticed without quite formalising it. Lateness at the start of a session is small and lateness at the end is large, not because the day gets worse but because delay accumulates while recovery is occasional. It is a ratchet, and the last slot inherits everything the earlier ones contributed.

Which is why the first slot is the best slot

The first appointment of a session begins from a clean position with nothing in front of it. It is the only slot in the whole day that can be guaranteed a start time, and the value of that has nothing to do with staff being fresher or systems being quieter. It is purely a consequence of position in the sequence.

The first appointment after a scheduled break carries most of the same advantage, because the break absorbs whatever overrun the morning produced. This is the practical reason for the well-worn advice to ask for the earliest slot or the first one after lunch. It isn’t a trick, and anybody working in the service will usually confirm it if asked directly.

Recovery only comes from the gaps that were built in

Well-designed schedules include slack: an unbooked slot, a deliberately short administrative gap, a break longer than it needs to be. Those exist precisely to absorb accumulated overrun and reset the session, and they are the first thing sacrificed when demand rises.

Removing slack raises utilisation and destroys punctuality at the same time, which is a genuinely difficult trade rather than a mistake. A session booked at full capacity uses every minute of the resource and will run late by design, since the first overrun has nowhere to go. A session with slack finishes on time and leaves visible idle periods that look, to anyone waiting, like waste. Both criticisms are fair, and no schedule can satisfy them together.

Non-attendance is what makes running late survivable

The main thing pulling a session back towards its timetable is people who do not arrive. That is an uncomfortable dependency and it’s real: a session booked to capacity typically only finishes near its planned time because a proportion of the bookings evaporate.

It creates a visible oddity worth explaining. A waiting room can empty suddenly and completely, several appointments can be called in quick succession, and the session can end almost on time after running forty minutes behind. Nothing was accelerated. A gap appeared in the sequence, and the ratchet finally released.

Two clocks, and they measure different things

The time on your letter is a scheduled start. The time you are actually seen is an outcome of everything upstream. Services that publish punctuality statistics usually report the gap between those two, averaged across a session or a week, and the average conceals the pattern entirely: the early appointments pull it down while the late ones carry the delay.

This is why the average waiting figure quoted at a service and your own experience can both be accurate and completely fail to describe each other. If your appointments have consistently landed late in sessions, you have been sampling the worst end of a distribution the average was designed to smooth.

Using the pattern rather than complaining about it

Ask what position in the session a slot occupies, not just what time it is. First of session, first after a break, or early in an afternoon are all materially different from mid-morning or last of the day, and booking staff generally know it. Where a service offers a choice, that is the axis worth choosing on.

And when a session is already running late, the honest thing to ask is how far behind it is, since that figure is knowable and is usually tracked. How long until I am seen is a harder question involving people whose appointments have not started yet. How late is the session is a fact somebody at the desk can read off a screen.