A busy door and a good night are different things, and the gap between them is where most venues quietly lose money. Here is what the headcount hides, and the four numbers that tell you whether a full room actually paid.

Friday felt busy. The door counted 140. The room looked full all night.

Then the takings come in and they are barely ahead of a Wednesday that counted 60.

This happens often enough that most owners stop trusting the door count — and then stop counting anything. That is the wrong correction.

The short version

A headcount tells you how many arrived, not how many spent
The number that matters is spend per head, split by how they came in
A full room with low spend per head costs more to run than a half room with high spend
Free and discounted entries can make a good night look like a bad one, and the reverse
Count the door, but never quote it without the second number

What the headcount hides

The door says	What it does not say
140 through the door	How many bought anything past the first drink
Busiest night this month	How many came in on a guest list
Room was full from 10pm	How long the average group stayed
Up 30% on last Friday	Whether the extra 30% spent anything

⚠️ Headcount is an input. Revenue is an output. Treating the input as the result is how a venue ends up busier and poorer at the same time.

One — spend per head, split by entry type

One number fixes most of this.

Spend per head = takings ÷ people through the door

But the split is where the answer lives:

Full price entry  …spend per head
Guest list        …spend per head
Comp / promoter   …spend per head

Friday …140 in, 700,000 taken ⇒ 5,000 per head
   full price 60 …450,000 ⇒ 7,500 per head
   guest list 50 …200,000 ⇒ 4,000 per head
   comp       30 …50,000  ⇒ 1,667 per head

⇒ The comp group filled the room and contributed almost nothing. The night looked busy because of them.

The question this answers

Ask: if I had let in 30 fewer comps, would I have taken less?

⚠️ Usually not — and you would have spent less on staff, security and stock to serve them.

Two — a full room is not free to run

Every extra body costs something even if they buy nothing.

Bar staff needed  …scales with the room
Security          …scales with the room
Cleaning, stock, breakages …scale with the room
Seats occupied by non-spenders …block spenders

⚠️ The last one is the expensive one and the hardest to see. A comp taking a table at 10pm means the group who would have taken it at 10:30 stood at the bar and left early.

How to check it

On your busiest nights, how many groups left within 30 minutes of arriving?

⇒ If that number spikes on the nights with the biggest door count, your room is full of people who are blocking the people you want.

Three — the promoter arithmetic nobody runs

Promoters are paid for heads. You are paid for spend. Those are not the same job.

Promoter brings 40
   you pay 800 a head            ⇒ 32,000 out
   those 40 spend 3,000 each     ⇒ 120,000 in
   ⇒ 88,000 gain, before cost of serving them

Now the same deal with a different crowd:

Promoter brings 40
   you pay 800 a head            ⇒ 32,000 out
   those 40 spend 900 each       ⇒ 36,000 in
   ⇒ 4,000 gain, and they filled forty seats all night

⇒ Same promoter, same fee, same headcount. One deal works and one does not.

⚠️ You cannot tell these apart without tagging entries by how they came in. Without the tag, both look like "a busy Friday".

What to change

Pay per head, but review on spend per head
Set a floor: if the promoter's crowd spends under X per head for two months, renegotiate

Four — the drink that measures the room

There is a quicker proxy if you cannot tag entries yet.

Second drinks ÷ first drinks

A room full of people who buy one drink and stand is a different room from one where most buy a second.

Healthy …most of the room comes back for a second
Weak    …a big first-drink number and a small second

⚠️ A large gap between first and second drinks usually means the room is full of people who came for the door, not the night.

Five — the night-of-week trap

Comparing Friday to Friday is right. Comparing Friday to Wednesday is not.

Wednesday …60 in, 300,000 ⇒ 5,000 per head
Friday    …140 in, 700,000 ⇒ 5,000 per head

Identical spend per head. But:

Friday cost more to staff, more security, more stock
⇒ Friday was the worse night per unit of effort

⇒ Compare each night against the same night, four weeks back. Anything else mixes two different businesses.

Six — what to do with the count you already have

Do not stop counting. Add one column.

Date ／ night ／ heads ／ takings ／ spend per head ／ entry type split

⇒ The split is the work. Everything else you already have.

If you can only add one thing

Tag every entry as full price / list / comp

⇒ That single tag turns a meaningless headcount into a usable one.

The numbers to keep

One   heads, split by entry type
Two   takings, and spend per head for each type
Three groups leaving within 30 minutes, on your busiest nights
Four  second drinks as a share of first drinks

⚠️ Three is the one that tells you the room is blocked rather than quiet. It reads as a service problem, and it is usually a door problem.

A sheet for the office

────────────────────────────────────
  Was it actually a good night?

  Heads …………   Takings …………
  Spend per head …………

  Full price …… heads, …… per head
  Guest list …… heads, …… per head
  Comp       …… heads, …… per head

  Never quote the door count on its own.
  Compare to the SAME night four weeks back.
────────────────────────────────────

Common objections

"A full room brings more people in"

⇒ Sometimes true early. Check whether it still holds after 11pm, when the people it attracts are the ones already inside not spending.

"We can't tag entries, it slows the door"

⚠️ Three buttons on the door is seconds a night. Not tagging costs you the ability to price every promoter deal you sign.

"Our promoter has been with us for years"

⇒ Then you have years of data. Run the spend per head for his crowd against everyone else's for two months.

"Spend per head is low because we're a late venue"

⇒ Possibly. Compare your own late nights against each other, not against a benchmark. The question is whether this Friday beat the last four.

"The comps are industry people, they're good for the room"

⚠️ Some are. Count how many and decide the number deliberately, rather than letting it drift with who is on the door that night.

This week

One   add an entry-type tag at the door: full price / list / comp
Two   for the next two Fridays, record heads by type and takings
Three work out spend per head for each type
Four  compare against the same Friday four weeks back
Five  if comp spend per head is under a quarter of full price, cap the number

Doing it by hand, or on a screen

Two Fridays by hand is easy. Every night, by type, compared to four weeks back, is not.

By hand …door sheet plus till total, split by type, month by month. Hours
On screen …heads by entry type against takings, with the same night four
           weeks back already lined up next to it

⇒ tasteck holds takings by night against customer records, so spend per head can be split by how the customer came in rather than averaged across everybody.

⚠️ It only works if the door tag reaches the system. A tag written on a clipboard and never entered leaves you exactly where you started.

Wrapping up

Headcount is an input, not a result
The usable number is spend per head, split by entry type
A full room costs more to run, and non-spenders block spenders
Promoter deals should be paid on heads but reviewed on spend
Compare each night to the same night four weeks back
If you add one thing this month, add the entry-type tag

A busy door and a good night are different claims. Only one of them shows up in the banking.

Door Count Is Not Revenue