Your Best Night Was Not Your Best Night

Saturday did fourteen thousand. Tuesday did three.

Everyone knows which one was the good night. Everyone might be wrong.

What has to come off first

A night's take is not a night's contribution. Four things sit between them, and all four are larger on the busy night.

Staffing for the night, not the week. Saturday ran fourteen people, Tuesday ran four. Until that is subtracted per night rather than averaged across the week, every busy night looks better than it was.

Comps and promotional entry. Concentrated on the nights you promoted. The cost lands on the same night as the take, and only one of the two gets recorded.

Cost of goods on what actually sold. Saturday's mix and Tuesday's mix are different — a night heavy on premium bottles and a night heavy on standard drinks have very different margins on the same revenue.

Security, door, and anything else that scales with the crowd. Fixed for the venue, variable for the night.

What usually happens when you subtract them

Two findings recur.

Your busiest night is still your best — but by much less than the take suggested. Often a third of the apparent gap survives. That matters, because decisions get made on the size of the gap: which nights to promote, which to consider closing, where to add staff.

One night you thought was fine is not. Usually a mid-week night carrying weekend-level staffing out of habit, or a promoted night whose promotion costs more than it brings. This is the finding that pays for the exercise.

The measurement

You do not need a costing system. One month, four numbers per night:

take
staff cost for that night
comps and promotional entry given that night
cost of goods on what sold that night

Subtract, and rank the nights. That is the whole method.

⚠️ Rank them, do not just look at the totals. The value is in the ordering changing — a night that moves from second to sixth is telling you something a percentage never will.

The trap in acting on it

The instinct after seeing the ranking is to cut the worst night.

Careful. Three reasons a weak night can be worth keeping:

It carries fixed costs. Rent and licences do not stop on the nights you close. A night contributing anything above its own variable cost is helping.

It holds the staff. People need enough hours to stay. Cutting a night can cost you the team that works the good ones.

It feeds the good night. Some guests discover you on a quiet Wednesday and come back with eight people on a Saturday. If nothing links a guest's first visit to their later ones, that contribution is invisible and the night looks worse than it is.

The right first move is almost never closing. It is fixing the staffing level and the promotional spend on that night, which is cheap and reversible.

The one number that reorders everything

Staff cost as a share of that night's take.

Most venues know the weekly figure and not the nightly one. The weekly figure averages a lean Saturday with a heavy Tuesday into something that looks acceptable and hides both.

⚠️ When you look per night, the outlier is usually not the busiest one. It is a quiet night rostered as though it were busy — and it has probably been that way for a long time, because nothing was looking.

Subtract the things that will not happen again

The one-off group  a party that booked once
The event    that will not recur
The weather   or a neighbour being closed

⚠️ What remains is the number worth planning from, and it is always smaller than the headline.

---

The night you celebrated was not the peak

The best night  turns out to be an ordinary night with one large booking
An ordinary night  turns out to have been the actual peak

⚠️ The night worth studying is frequently not the night you celebrated. 　 ⇒ And it went unexamined, because nothing about it stood out.

---

The measurement that survives an argument

Not  total takings
But  takings excluding anything that cannot recur

⚠️ Anyone can dispute a total. Nobody disputes a list of exclusions, because the exclusions are individually obvious.

---

Repeating the peak means chasing the one-off

Trying to repeat the peak  usually means chasing the one-off

⚠️ The large booking was not caused by anything you did, so repeating your own actions reproduces the ordinary night underneath it. 　 ⇒ Which is the correct thing to repeat, and it looks like a 　 　 disappointment.

---

Capacity-limited or demand-limited

Turned away

⚠️ A night with high takings and refusals was capacity-limited; a night with high takings and none was demand-limited, and those call for opposite decisions. 　 ⇒ Without it, both nights look identical in the report and get the same 　 　 response.

---

---

Three to hold

Contribution per night, after the four subtractions. The ranking.

Staff cost as a share of take, per night. The number that identifies the rota to change.

Comps as a share of take, per night. Concentrates where you promote, and is almost never counted against the night that received it.

Where the record has to sit

Take is in the till, staff cost is in the rota, comps are in someone's head, and cost of goods is on an invoice. Four systems, so the subtraction never happens, and nights get ranked by the only number that is easy to get.

tasteck keeps shifts, sales, and comps against the same nights, so contribution per night is a report rather than a month of reconciliation.

The take tells you how busy you were. It does not tell you what the night was worth.

---

Read next

• Comps and House Accounts: What They Actually Cost
• The Slow Tuesday You Keep Open
• Night Venue Scheduling: Why the Roster Never Matches the Room
• The Slow Tuesday Problem
• Why Your Best Staff Leave in Month Four
• The Comps You Never Approved