The Number That Went Up

Takings were up eleven percent last month. Everyone was pleased, nobody asked why, and the following month it was back to normal.

Something caused that eleven percent. Nobody knows what it was, so nobody can do it again.

Why good news never gets examined

Bad numbers demand an explanation. Good ones do not, so none is produced.

Asking feels ungrateful. ⚠️ Interrogating a good month reads as doubting the team, so the question does not get asked out loud.

The explanation is assumed. Usually attributed to whatever was recently changed, whether or not it was the cause.

There is no process for it. Every venue has some version of "why was that bad". Almost none has "why was that good."

The four candidate explanations

Nearly every good month is one of these, and they have completely different implications:

Something you did. The thing worth knowing about — repeatable, and the whole reason for asking.

Something in the calendar. An extra weekend night, a holiday landing well, a local event. ⚠️ Entirely predictable and entirely outside your control — and it will not repeat next month.

One large transaction. A single party or booking that moved the total. Check the median as well as the total — if the median did not move, one guest did this.

A measurement change. ⚠️ The one nobody suspects. A different date range, a category moved, a fee recorded differently. It is more common than people expect and it produces a rise that never existed.

How to ask without spoiling it

Ask about the mechanism, not the credit. "What do we think drove that" is a different question from "who is responsible."

Ask the floor first. ⚠️ They frequently know exactly what happened — a group that came three times, a night that went unusually well — and it does not appear in any report.

Say why you are asking. "If we did something that worked, I want to do it again" removes the entire suspicion that the question is a challenge.

The check that takes ten minutes

Compare the median as well as the total. Total up, median flat means one large item.

Compare the same trading days. ⚠️ Five Saturdays against four is not a comparison, and this alone explains a large share of month-on-month movement.

Check whether the definition changed. Ask whoever produces the number whether anything about how it is produced has changed.

Then look at what you actually did. Only after ruling out the first three, because otherwise you will credit yourself for a calendar.

What happens when you skip it

The wrong thing gets repeated. Credit is given to a recent change that was not the cause, and it becomes established practice on false evidence.

The right thing gets dropped. ⚠️ The actual cause was never identified, so it was never protected — and it can be removed by accident.

The next dip is unexplainable. With no understanding of the rise, the fall looks like a new problem rather than a return to normal.

The same applies to a good night

Not just months. A night that goes unusually well is worth ten seconds of questioning while everyone still remembers it — by next week the detail is gone, and the detail is where the cause lives.

Three to hold

Total and median together. ⚠️ Always both. One of them being up alone tells you which kind of month it was.

Trading days in the period. The most common false explanation, and the easiest to rule out.

Whether anyone was asked why. A good month with no stated explanation should be treated as unexplained, not as success.

Where the record has to sit

Answering "why was that good" a month later needs the detail still to exist — which bookings, which nights, which guests. If the record is a monthly total, the question is permanently unanswerable and every good month stays a mystery.

tasteck keeps takings against the bookings and guests that produced them, so a month can be opened up rather than only compared.

You did not have a good month. Something specific happened, and it is still worth finding out what.