A mid-sized restaurant with three managers on a rota. Reviews across one quarter kept mentioning the wait for the card machine at the end of the meal — eleven times in thirteen weeks.
It was raised at pre-shift four separate times, by three different managers, as a new observation. Each time, the same remedy was proposed: remind the team to take the machine to the table proactively. Each time it improved for about a week, which is roughly the length of one manager's run of shifts.
Nothing was wrong with any of the four responses. The problem was that each manager was starting from zero, because the only record of the previous three attempts was in somebody else's head, on their days off.
What a rota does to institutional memory
A single-manager venue has continuity by accident. The person who read the review on Monday is the person standing at the pass on Friday, so the knowledge travels for free.
Add a second and a third manager and that stops being true in a way that is easy to miss, because everyone is doing their job. The specific things that get lost:
That it was raised before. Each manager experiences the eleventh card-machine review as the first one they have seen.
What was already tried. The reminder was issued four times. Nobody knew it had failed three times, so nobody concluded that reminders were not the answer.
Why the obvious fix was rejected. One manager had worked out that the machines were kept at the host stand, five tables away, and had asked for a second holster near the back section. It was declined on cost in week three. Managers two and three never knew the question had been asked, so they did not ask again.
Who the guest was. A regular who complained twice gets handled as a first-time complainer twice.
The cost here is not the eleven reviews. It is a quarter spent applying the same failed remedy, with sincerity, four times over.
The brief as memory rather than report
Most weekly brief templates are designed to send information upward — a summary for an owner or an area manager. That is a legitimate use and it is not the scarce one.
The scarce thing in a multi-manager venue is a document that answers, for whoever is on tonight: what are we currently working on, what have we already tried, and what did it do?
Which means the brief needs two things almost no template includes: a line that carries forward, and a record of what has been ruled out.
The format
One page, in a shared place the whole management team can edit, written weekly. Five blocks.
1. Open items, carried forward
The most important block and the one that gets dropped because it requires last week's document.
| Theme | First seen | Tried so far | Current state |
|---|---|---|---|
| Card machine wait | Week 31 | Pre-shift reminder ×3 (holds ~1wk); 2nd holster declined wk 33 | Re-costing holster; reminder not working |
| Terrace drinks wait | Week 36 | Dedicated runner Fri/Sat from wk 38 | Watching — 2 wks data |
Nothing in that table is clever. The Tried so far column is the entire point: it prevents the fourth reminder, and it makes the holster request arguable, because a declined £180 holster against a quarter of failed reminders is a different proposal from a £180 holster against nothing.
2. New this week
Only what has not appeared before. Two or three lines. If it is already in block 1, it does not go here — it updates the state column there instead.
3. One action, one name, one date
Singular. A brief listing six actions produces none. Four lines is genuinely enough for a weekly summary and the same restraint applies here.
4. Ruled out
A standing list, appended to and never cleared. Reminders do not fix the card machine. The cold-food complaints were not the kitchen — it was the route from pass to the back section. Noise complaints are the hard-floor section only, not the whole room.
This block is the cheapest thing in the document and the most valuable after a year. It is the accumulated negative knowledge of your management team, which otherwise evaporates every time somebody leaves.
5. What to say at pre-shift
One sentence, written out, so that three managers deliver the same message rather than three variations of it. Consistency matters more than phrasing here — a team hearing the same sentence from whoever is on duty treats it as a standard, and three different versions as three opinions. How that sentence lands is worth getting right.
The rules that keep it alive
It is edited, not rewritten. Each week's version starts as last week's with the state column updated. Writing it from scratch guarantees block 1 and block 4 disappear, which removes the only reason it exists.
Whoever is on Monday writes it. Not a fixed person, which creates a single point of failure and a document owned by someone who is not always there. Rotating authorship also forces each manager to read the open items properly, which is most of the value transferring.
It lives where all three can edit it. A shared document, not an email. An email is a snapshot; this needs to be a current state.
Fifteen minutes, hard stop. Four blocks of it are updates to existing lines.
How this fails
It can become a complaints archive. Block 1 grows forever if nothing is ever closed. Close an item when the theme has been absent for four weeks, and move the finding to block 4. A brief with nineteen open items is a document nobody reads.
Rotating authorship produces inconsistent quality. Accept it. A mediocre entry in the right place beats an excellent one in a manager's notebook.
It does not replace reading the reviews. The brief is a summary and summaries destroy the specificity that makes a review useful — the guest who said the chicken was dry the second time has told you something a theme count cannot hold. Somebody still has to read the text. And where that reading sits in the organisation determines whether anything changes.
One site, one document. For several sites this is the wrong instrument — the comparison question is different and needs rates against peers rather than a shared memory. That is a reporting problem with its own escalation structure.
It will feel unnecessary for about six weeks. The payoff arrives the first time a manager returns from a week off, reads block 1, and does not re-propose something that has already failed three times. That is not a satisfying week-one return, which is why most versions of this get abandoned in week two.
The card machine, incidentally, was fixed by the £180 holster. It took a quarter to get there, not because anybody was slow, but because the argument for it had to be made four times by people who did not know the other three had made it.
OMMU keeps the theme history and what was tried against each one, so block 1 is populated rather than remembered. A shared document and a Monday habit does exactly the same job — the part that cannot be skipped is editing last week's rather than starting a new one.