Skip to content
Queued

31 July 2026 · 4 min read

Quoting wait times accurately: the maths behind a number people trust

How to calculate a wait estimate guests actually believe, why round numbers erode trust, and what to do when the queue speeds up or slows down.

"About 20 minutes" said to everyone, all night, regardless of how long the line actually is, is the single most common reason guests stop trusting a wait estimate. It's understandable — staff are busy and a round number is fast to say — but it trains people to ignore the number entirely, which is worse than not giving one at all.

The basic formula

Most walk-in wait estimates come down to a version of the same calculation:

Three tables free and a five-minute average service time means someone six people back should expect roughly ten minutes, not the twenty a tired host might say by habit. The maths isn't complicated — the hard part is keeping the average service time honest and up to date.

Why round numbers fail

  • They don't move. If the same "20 minutes" is quoted to the 2nd person in line and the 20th, whoever's actually further back learns fast that the number is fiction.
  • They don't reflect reality mid-shift. A slow start and a fast finish average out to a number that was wrong for most of the night.
  • They erode the whole system's credibility. Once a guest catches one wrong estimate, they stop trusting the next one — even if it's accurate.

Getting the estimate right

Use a rolling average, not a fixed setting

Set a starting average service time based on your normal pace, but let the actual completed tickets from the last hour (or the last N tickets) pull that number up or down as the shift goes. A queue that's moving faster than usual should show shorter waits without anyone touching a setting.

Account for concurrent capacity, not just a queue length

Position alone isn't enough — being 8th in line means something very different with three tables turning over than with one. Estimates that only count position and ignore capacity are consistently wrong in venues with more than one server, chair, or counter.

Show a range or round to something honest

"15–20 minutes" sets more realistic expectations than a precise-looking "17 minutes" that implies false confidence, or a suspiciously round "15 minutes" that sounds guessed. Pick whichever convention matches how confident the estimate actually is.

Update it as things change

A guest who joined at position 12 with a 30-minute estimate, and is still shown 30 minutes ten minutes later despite the queue moving, will assume the number was never real. The estimate on their status page should recalculate as the queue moves, not freeze at the value it showed when they joined.

2xroughly how much more likely guests are to wait out a queue when the estimate visibly moves, versus a static number shown once

Handling the edges

SituationWhat to do
Sudden rush (large party, event let-out)Let the estimate spike honestly rather than hiding it — a temporarily long wait shown accurately is better than a short one that turns out false
Very early in service, few data pointsFall back to a sensible manual default until enough tickets have completed to calculate a real average
A queue that's effectively stalled (kitchen issue, staff shortage)Some tools let you pause the queue or show an announcement rather than let the estimate quietly balloon with no explanation
Multiple party sizes with different service timesWeight the average by party size if large groups genuinely take longer — a flat average under-quotes big tables and over-quotes small ones

The trust payoff

None of this is about being clever with maths — it's about not lying to people, even by accident through laziness. A venue that consistently quotes honest, moving estimates gets guests who wait out the line instead of leaving, and a host stand that isn't fielding "how much longer?" every few minutes, because the guest already has the answer on their phone.

Where Queued fits

Queued calculates the wait estimate from live position, your concurrent service setting, and a rolling average of recent ticket completion times — not a number typed in once and forgotten. You can also switch a queue to a manual estimate if you'd rather set it yourself for a specific event.

Give guests a number they can trust

Queued recalculates wait estimates live as your queue moves.

Get started free
Should wait estimates be exact numbers or ranges?
A range (e.g. "15–20 minutes") is usually more honest than a precise-looking single number, especially early in service before enough data has built up.
How is estimated wait time actually calculated?
Typically: people ahead of you, divided by how many parties can be served at once, multiplied by an average service time — refreshed from recent completed tickets rather than a fixed guess.
What if the queue suddenly gets much busier?
Let the estimate rise and show it honestly. A temporary long wait shown accurately keeps trust; a short estimate that turns out wrong loses it.

Written by Queued · Operations

Start this afternoon

Stop keeping the list on a napkin.

Set up a queue, print the QR, and let the next person wait in the pub across the road instead of your entrance. Free plan, no card.