Events
How to do an accurate headcount at a door or event
An accurate headcount needs three things: one counter per entrance, a separate count for people leaving, and a fixed rule for what counts as a person. Count on the same physical trigger every time, such as a foot crossing the threshold, and record the total at fixed intervals so a single error cannot corrupt the whole night.
What actually causes miscounts
Headcount errors are rarely a failure of arithmetic. In practice they come from four places, and each has a different fix:
- Bunching. Three people come through shoulder to shoulder and register as one or two. This is the single largest source of error at a busy door.
- Ambiguity. Staff, children in arms, someone stepping out for a phone call and coming back. Without a rule agreed in advance, two counters on the same door will disagree.
- Attention loss. Anyone counting for more than about twenty minutes without a break drifts, especially if the flow is irregular.
- Irrecoverable slips. With a mechanical clicker, a double press cannot be taken back, so counters either accept a known error or reset and lose everything.
Set the counting rule before anyone arrives
Write down, and tell every counter, exactly what triggers a count. A good rule is physical, unambiguous and identical for everyone:
- The trigger: a person's leading foot crossing the threshold line. Not eye contact, not ticket scan, not “when they get past me”.
- Who counts: decide in advance about staff, performers, contractors and infants. For fire-safety occupancy every body counts, including staff.
- Re-entry: if people can leave and return, you must run an out count, or your in-count will inflate through the evening.
For bunching, the practical fix is to count the trigger rather than the person: watch the threshold line, not the faces. If groups are consistently large, station a second counter and split by side.
In and out counts for live occupancy
A single count answers “how many arrived tonight”. Fire safety, licensing and capacity management all need a different number: how many are inside right now. That requires two counts running together.
- Run one counter named In and one named Out.
- Set the maximum on the In counter to your licensed capacity, so it warns you rather than silently passing it.
- Occupancy is In minus Out. Recalculate it out loud at fixed intervals, not only when someone asks.
Handheld clickers make this awkward, because it means carrying two of them and remembering which thumb is which. An on-screen counter keeps both totals visible in one list, which removes that error entirely.
Splitting a large crowd between counters
Above roughly 400 people, or more than one entrance, a single counter is not reliable. Split the job on a fixed boundary rather than by judgement:
| Situation | How to split | Watch out for |
|---|---|---|
| Several entrances | One counter per door, named after the door | Doors opened late or closed early without telling the counters |
| One wide entrance | Two counters, split on a physical line such as a barrier or pillar | People drifting across the line mid-entry |
| Seated audience | Count by block or row range, one counter each | Empty reserved seats counted as occupied |
| Standing crowd | Count a marked sample area, then multiply by area ratio | Density is never uniform, so use several sample areas |
Give each counter their own device with a counter named after their zone. Nothing needs to sync: at the end, each person reads their total aloud or copies it into a message, and one person adds them up.
Checking your total
A headcount you cannot check is a headcount you cannot defend. Three cheap checks:
- Interval snapshots. Write the running total every 15 or 30 minutes. If one interval is wildly out of line with the rest, you know roughly when the error happened.
- A second independent count. Two counters on the same door, counting separately. Agreement within about 2% is good; a wider gap means your counting rule is ambiguous.
- Cross-check against another source. Tickets scanned, wristbands issued, or bar transactions. These will never match exactly, but a large divergence is a signal worth chasing.
Rotate counters every twenty minutes or so on a busy door, and record the total at each handover so a lapse in attention costs you one interval rather than the whole evening.
Frequently asked questions
How do you count people accurately at a door?
Set one unambiguous physical trigger, such as a leading foot crossing the threshold, and count that rather than judging each person. Run a separate out count if people can re-enter, and write down the running total every 15 to 30 minutes.
How do you work out how many people are inside a venue?
Run two counters, one for people entering and one for people leaving, then subtract the out total from the in total. A single arrivals count tells you attendance for the night, not how many are inside at this moment.
What is the most common headcount mistake?
Bunching, where several people cross the threshold shoulder to shoulder and register as one. Watching the threshold line rather than people's faces helps, and a second counter splitting a wide entrance on a fixed physical boundary helps more.
Can several people count the same event on different phones?
Yes. Give each counter their own device with a counter named after their door or zone, then add the totals at the end. Nothing syncs between devices, so no attendee data ever leaves the phone it was counted on.