How to run a workshop sign-up with capacity limits
Decide first whether seats go to whoever signs up first or to whoever wants them most. First come is quick and needs no tool. For ranked choice, set a closing date, collect rankings, place everyone at once, and check the room sizes before you tell anyone, because the rooms can matter more than the method.
By Sortround. Published 5 November 2026.
First come or ranked choice?
| What matters | First come | Ranked choice |
|---|---|---|
| People know their place | The moment they sign up | After the closing date |
| Your work | Almost none | Set a date, then place and publish |
| Who gets the popular seat | Whoever saw the message first | Decided from everyone's choices together |
| Good for | Drop-in sessions, spare seats, a sign-up that stays open | Popular workshops, people on shifts, anyone who cannot be at a keyboard at nine |
If every workshop has more seats than interest, use first come and stop reading. When seats are tight it leaves the problem raised in a teacher's blog post from May 2016: nothing makes sure each person gets one of the sessions they wanted most.
The manual method
First come. Use a form with one question per person, and take a workshop off the list when it fills, by hand or with a form add-on that hides a full option. Keep a waiting list in the same sheet. The slip to watch for is two people taking the last seat at almost the same moment.
Ranked by hand. Collect three choices per person, count the demand, draw for the full workshops, then move those left to their next choice. The electives guide has the steps and the spreadsheet formulas.
Either way, count first. Before anyone is told, set seats against demand for each workshop. That sum is where most of the result is decided, as the example shows.
Worked example: the same 48 people, three room plans
A training day with 5 workshops running at the same time. Each of the 48 people ranks 3, and the rankings never change below: only the rooms do. Difficult conversations is the popular one: 20 put it first. Nobody here is real, but every count is what the site's solver returns for these rankings.
| Workshop | Put it first | Plan 1 seats | Plan 2 seats | Plan 3 seats |
|---|---|---|---|---|
| Difficult conversations | 20 | 8 | 16 | 16 |
| First aid | 6 | 16 | 8 | 8 |
| Public speaking | 6 | 8 | 8 | 8 |
| Spreadsheets | 12 | 8 | 8 | 8 |
| Project planning | 4 | 12 | 12 | 6 |
| Plan | Seats | 1st | 2nd | 3rd | Unranked | No place |
|---|---|---|---|---|---|---|
| 1. Rooms as first booked | 52 | 32 | 11 | 1 | 4 | 0 |
| 2. Big room swapped to the popular workshop | 52 | 40 | 8 | 0 | 0 | 0 |
| 3. Swapped, and one room loses 6 seats | 46 | 40 | 6 | 0 | 0 | 2 |
Plans 1 and 2 have the same 52 seats. Moving Difficult conversations into the big room takes first choices from 32 to 40, and the people placed in a workshop they did not rank from 4 to 0. No method could reach 40 first choices under plan 1: the ceiling there is 32. Solve's order also took 2 more first choices at the cost of 4 people outside their lists; a method that put lists first could have had 30, 13, 5 and nobody unranked. On this round, switching off placing people in sessions they did not rank gives exactly that.
Plan 3 shows the limit itself. With 46 seats for 48 people, 2 people have no place, and the first choices do not change. Solve places as many as it can and lists the rest as No place yet, for you to place by hand or to fix by finding seats.
Here is the same day run first come, as if people arrived in 500 different orders and each took the best of their 3 that still had a seat. Under plan 2 that gives this range:
| Method | 1st | 2nd | 3rd | None of 3 |
|---|---|---|---|---|
| Solve | 40 | 8 | 0 | 0 |
| First come | 36 to 40 | 2 to 10 | 0 to 8 | 0 to 3 |
On a good day first come gets close: it matched Solve in 2 of 500 orders. On a bad one, 3 people miss all three, and which people depends on when they read the message.
Doing it on Sortround
- Add the workshops with their capacities, and set a closing time.
- Share one event link. People rank by dragging, with no account and no email. In adult mode they can join by the link with their name and get a code.
- When choices close, press Solve and read the fairness view.
- Change a capacity and solve again, as many times as you like, before you publish. That is the room-plan table above, on your own numbers.
- Decide what happens when someone misses all their choices. By default they go to a workshop with a seat, and you can switch that off so they show as No place yet.
- Publish. Each person sees only their own place. A late joiner takes a free seat on the next solve, and nobody already told is moved.
This example fits without paying: 60 people and 20 sessions in 1 round. More people, a second round of the same workshops, or moving someone already told needs a paid plan.
Pages read for this guide, as they stood on 5 November 2026: a teacher's blog post (May 2016).