The Week Before The Big Night

You have known about it for months. It is the biggest night of the year and it still goes wrong in the same three places.

A known peak is not a capacity problem. It is a preparation problem, and preparation has a schedule.

The three failures

Every venue, every year:

Not enough of the right people. Not headcount — the specific people who can handle a section under pressure. Frequently the rota is full and the capability is not.

Running out of something. Almost never the main line. It is the second-tier item that sells at three times its usual rate because the crowd is different.

The room jamming at one point. The door, the bar, the cloakroom, the toilets. ⚠️ The bottleneck on a peak night is rarely the one that constrains a normal night, which is why it surprises people annually.

The schedule

Four weeks out — people. Confirm who is working, in writing. Everyone in the area is competing for the same staff and the good ones commit early. ⚠️ Leaving this to two weeks out is the single most common cause of the first failure.

Two weeks out — stock, against last year. Not against a normal week. Pull last year's actual consumption for that night — this is why the debrief matters, and if you have not got the data, this is the year to start recording it.

One week out — the flow. Walk the room and ask where three hundred people go. The bottleneck is usually visible in advance and cheap to fix with a table moved or a second position opened.

The day before — the decisions. What happens if the queue exceeds X. Who can authorise what. When last entry is. ⚠️ All of these will otherwise be decided at 23:00 by whoever is nearest, which is the definition of the night going wrong.

What not to do

Do not add capacity you have never run. A peak night is the worst possible time to try a new layout, a new system, or a new supplier. Whatever is untested will fail at the worst moment.

Do not schedule your best people across the whole night. They will be exhausted by the hour that matters. Stagger them so the strongest cover lands where the pressure is, not at 19:00.

Do not take bookings you cannot honour. The temptation is enormous and the cost lands on the night, in public, in front of everyone.

The debrief

Fifteen minutes, the following week, while it is still fresh.

Three questions:

• What ran out?
• Where did it jam?
• Who did we need more of?

Write the answers where next year's version will be read. ⚠️ This is the entire mechanism by which a peak night gets better — and it is skipped almost universally, because the week after is when everyone is recovering.

Fifteen minutes now saves the same three failures next year.

The number worth knowing

Consumption on the peak night, per line, against a normal night.

Not revenue — units. It tells you what the different crowd actually buys, and it is the only reliable basis for next year's order.

The night does not fail on the night

What happens on the night  the consequence
What caused it     something not decided the week before

⚠️ On the day you can only cope. The last day you can actually fix anything is several days earlier.

---

Three failures, and the invisible one

Not enough people  asked too late
Not enough stock  ordered on the usual cycle
Nobody deciding  everyone deciding, on the night

⚠️ The third gets overlooked and costs the most. A night with enough people and enough stock still collapses if four people each answer the same question differently.

---

Work the schedule backwards from what needs outsiders

Suppliers, extra staff, repairs  furthest back
Anything you can do yourselves  nearest the day

⚠️ Starting with what is visible uses up the days in which other people could still have been asked.

---

Do not add anything new on the night

Something new  fails in a way you have not seen before

⚠️ A big night is for doing the usual thing more intensely. 　 ⇒ New things get tried on quiet nights, where a novel failure costs 　 　 almost nothing.

And do not concentrate the decisions

⇒ ⚠️ If one person holds everything, the night stops when they do. Two decision-makers, split by area, costs nothing to arrange in advance and cannot be arranged during.

---

Debrief the next day, not the next week

The next day  you remember what actually happened
A week later  a clean account that helps nobody

⚠️ Write three things: what ran out, what hour it broke, and who was stuck unable to decide. 　 ⇒ Almost nobody writes the third. Stock and staffing get recorded; 　 　 the gaps in authority never do.

---

The number worth knowing is the refusals

Takings  tells you what you served
Turned away  tells you whether you were at the ceiling

⚠️ A large refusal count means the next one can be bigger. 　 ⇒ Revenue alone cannot tell you whether you hit a limit or simply met 　 　 demand, and those call for opposite decisions.

---

---

Three to hold

Staff confirmed four weeks out, as a share of the rota. The leading indicator for failure one.

Items that ran out, recorded on the night. ⚠️ If it is not written down that night, it will not be remembered in December.

Debriefs held, as a share of peak nights. Almost always zero, and it is the cheapest improvement available.

Where the record has to sit

Last year's consumption, last year's staffing, and last year's problems live in three places or in nobody's memory. That is why the same three failures repeat — not because they are hard, but because the evidence from last time was never kept.

tasteck keeps sales and shifts against dates, so last year's version of the night is a report rather than a recollection.

You have run this night before. The only question is whether last year left you anything to work from.

---

Read next

• The Season You Hire For Too Late
• What The Weather Does To Your Week
• Night Venue Scheduling: Why the Roster Never Matches the Room
• The Slow Tuesday Problem
• Why Your Best Staff Leave in Month Four
• The Friday That Was Not Actually Good