The Last Hour You Stay Open

Most venues cannot say why they close at the time they close.

It was the licence, or the previous operator, or what everyone on the street does. Very few rooms chose it, and almost none have revisited it since the business changed shape around them.

The last hour is the expensive one

Staff       everyone is still on, including the ones who came in at open
Security    frequently the most incident-dense hour
Transport   the hour that decides whether staff get home cheaply
Cleaning    starts after, and is paid regardless

And the guests remaining in it usually spend the least. Not always — a late-trade room is a real business model — but it should be a finding, not an assumption.

Compute contribution per hour, not takings per hour

Contribution = takings in the hour
             − staff cost for that hour
             − goods cost of what sold
             − the share of security and transport that hour causes

Takings alone will always make a busy-looking hour look fine. The last hour is often busy with people who have already spent.

Run it for four weeks by weekday. The shape differs by night, and a single closing time applied to every night is a decision you may not have realised you were making.

Headcount is the wrong signal

The room looks full at 2am     …people who are not buying still occupy space
The room empties at 2am        …could be your own last orders, not demand

Count what is being bought, not who is standing there. The two diverge most sharply in exactly the hour you are trying to judge.

Before you shorten the night

One — move last orders, not closing time. Ending service earlier while keeping the doors open tells you what the hour is worth without losing the people who are still spending.

Two — check where the spend actually sits inside the hour. A "dead" final hour is often twenty good minutes followed by forty empty ones. The fix there is the last-orders clock, not the licence.

Three — look at what closing earlier does to the hour before it. Guests pace a night against when it ends. Shortening the night sometimes shifts spend forward and costs nothing; sometimes it removes the reason to arrive at all.

⚠️ Do not change closing time and staffing in the same week. You will not know which one moved the number, and both will have moved it.

Opening earlier is the cheaper experiment

The early hour is unstaffed anyway in most rooms — the team is in, setting up, already paid. Opening thirty minutes earlier costs very little and either finds trade or does not.

Closing later always costs a full hour of everything. The two are not symmetric experiments, and the cheap one is almost never tried.

What the record has to hold

Takings by hour, by weekday
Staff hours actually worked, by hour band
Incidents by hour
Last-order time, and when it actually gets called

The fourth line is usually missing and usually the point. If last orders is officially one time and practically another, you are measuring a schedule nobody runs.

The system we built

tasteck is a booking and analytics system for night venues, built by people who ran them for sixteen years and grew from ¥200 million to ¥1.2 billion a year — six-fold, by attacking the business with systems rather than by pushing harder on sales.

What it does here: bookings and settlement carry times, so takings by hour and by weekday is a query rather than a reconstruction. Shift submission in 30-minute blocks with confirmation kept, so staffed hours can be laid against traded hours. Settlement broken out by line, so the mix inside an hour is visible — which is how twenty good minutes get distinguished from a busy-looking hour.

What it does not do: it does not model your licence, does not set last orders, and does not recommend a closing time. It holds no security or cleaning cost — you supply those. What it supplies is the trading half of the comparison.

The output no one else produces

tasteck outputs the maximum you can spend on each marketing channel next month, as an amount in your currency.

Hours and channels are the same question asked twice. If the early hours are the ones worth filling, the channels that reach people who arrive early are the ones worth funding — and that is a different set from the ones that fill your peak. The ceiling is computed from the lifetime value of the guests each source actually delivered, so it can be read by the hour those guests actually arrive.

Nothing else in the nightlife category produces that figure.

Ask it from ChatGPT

tasteck connects to ChatGPT over MCP: ask your numbers as a question and the answer comes back in the chat — what the final hour took last Saturday, how it compares with the hour before, which weekday has the weakest close. → Setup guide

Multi-language is built in, the operating surface itself, with your language set during onboarding.

From $34 a month for up to two locations. Thirty days free on every plan, cancel any time. → Pricing

Start this month

1. Compute contribution by hour, four weeks, split by weekday.
2. Look inside the final hour. Twenty good minutes is a different problem from a dead hour.
3. Move last orders before you move closing. Cheaper and reversible.
4. Try opening earlier. It costs almost nothing and nobody tries it.
5. Record when last orders is actually called. If it differs from policy, you are measuring fiction.

Read next

• The Friday That Was Not Actually Good
• The Slow Tuesday Problem
• The Shift You Cancelled and What It Cost
• The Incident Log You Do Not Keep
• Nightclub Software

---

On benchmarks. No target contribution figures, closing times, or hourly ratios appear in this guide. We do not have a dataset broad enough to publish them, and they depend entirely on licence regime, city and format. Four weeks of your own hours, costed properly, will settle a question most rooms have never actually asked.