Every restaurant has one. Five minutes before doors, everyone standing in a rough circle, a manager reading out the covers, the specials and a reminder about phones. It happens six days a week, it costs about half an hour of paid time across the week, and in most places it teaches the floor nothing at all.
That is a lot of recurring time producing no learning, and it is the cheapest thing in the building to fix.
Why the standard pre-shift produces nothing
Four problems, and they compound.
It is entirely one-directional. One person talks, fifteen people listen, and listening is passive. Passive exposure produces a feeling of familiarity rather than the ability to produce an answer at a table, which is the central finding of the research on retrieval practice.
It covers too much. Covers, specials, VIPs, the broken table, the new till procedure, a reminder about uniform. Working memory tops out at a handful of items, so a briefing with nine topics delivers roughly none of them.
It is delivered at the worst moment. Two minutes before doors, with people finishing side work and thinking about their section.
And nothing is ever checked. Nobody is asked to say anything back, so nobody discovers what did not land until a guest does.
| Standard pre-shift | What it produces |
|---|---|
| Manager reads the specials | Recognition, not recall |
| Nine topics in five minutes | None of them retained |
| No questions asked back | Gaps discovered at the table |
| Same format daily | Attention drifts within a week |
The one change that matters
Ask one question and wait for the answer.
Not a rhetorical question, not “any questions”, not a quiz. One specific question to one specific person, out loud, with the rest of the floor listening.
“Marta, describe the special.”
That is the whole intervention. It takes twenty seconds, it converts the meeting from broadcast to retrieval, and everyone else rehearses silently while they wait to see whether they would have got it. The rest of the floor answering in their heads is doing exactly the work that makes the information stick.
Rotate who gets asked. Once people know it is coming, they prepare, which moves the learning to before the meeting where it belongs.
What to cover, and what to cut
A pre-shift should carry three things and no more.
The one thing that changed today. A special, a shortage, a large booking, a new dish. One.
The one thing you want better tonight. A specific behaviour, not a value. “Check back within two minutes of the first bite” is usable; “give great service” is not.
The one question asked back. As above.
Everything else belongs somewhere other than a briefing. Covers and VIP notes go on a printed sheet. Procedural changes go in writing where people can reread them. Individual feedback happens individually.
Cutting a pre-shift from nine items to three is what makes the three land, and managers consistently underestimate how much this improves recall.
| Keep | Move elsewhere |
|---|---|
| The one change today | Cover numbers: printed sheet |
| One specific behaviour | Procedure changes: in writing |
| One question asked back | Individual feedback: one to one |
| Thirty seconds on an allergen | Long menu training: a tasting |
Building a week out of it
One question a day compounds in a way a single training session does not, because spaced repetition beats one long block at nearly everything.
Monday: describe the new special. Tuesday: which mains come without a side. Wednesday: name the allergens in the three most-ordered dishes. Thursday: what pairs with the fish. Friday: describe the special again.
Five questions, twenty seconds each, and by Friday the floor knows the menu change properly rather than half-knowing it. This is distributed practice applied to a meeting you were holding anyway.
Keep a list of what you have asked. Managers who improvise the question end up asking about the same two dishes for a month.
The allergen thirty seconds
Worth protecting as a fixed slot, because it is the one topic where a gap has consequences beyond service quality.
Rotate through the dishes rather than repeating the same warning. “Which of tonight’s specials has dairy in it” is a question; “remember to check allergens” is noise that people stop hearing within a week.
The point is not compliance theatre. UK food businesses are legally required to provide allergen information, and the floor is where that requirement actually gets met or missed. Thirty seconds a day, asked as a question, does more than an annual training module.
Pair it with the standing rule that “I will check” is always an acceptable answer, so that nobody guesses to avoid looking uninformed in front of the team.
Running it when the floor is split
Most restaurants cannot get everyone in a room. The format survives.
Run it twice, identically, at the two shift starts. The second run takes the same five minutes and prevents the half of the floor that missed it from learning the menu off the printed card.
Write the three items down where the late arrivals can read them, and ask the question of a late arrival when they get there.
Do not let the pre-shift become a thing only the openers get, which is the default and which produces a visibly two-speed floor by the weekend.
What managers get wrong
Using it for discipline. A briefing that sometimes contains a public telling-off becomes a meeting people brace for, and braced people do not learn.
Talking for twelve minutes because the chef came out. Cap it, genuinely, and take the extra content to a proper tasting instead.
Asking “does everyone understand”, which invites a nod and confirms nothing.
Reading the specials from the sheet rather than having the floor describe them, which is the passive version of the exact thing you are trying to fix.
And skipping it when busy, which is when it matters most, because a busy service is when knowledge gaps become visible to guests.
What it looks like after a month
Servers arriving already knowing tonight’s special, because they know they might be asked.
New starters getting daily low-stakes practice instead of waiting for a training day that keeps being postponed, which shortens the first week considerably.
A floor that can describe dishes in one clean sentence rather than reading the menu back.
And a manager who knows where the gaps are, because they have been asking questions rather than making statements.
None of that costs additional time. It is the same five minutes, spent differently.
A script you can use tomorrow
The format is easier to adopt when it is written down, because the temptation to add a fourth item is constant.
Open with the change. “One thing tonight: the bream is off, the special is a monkfish, it comes with the same braised fennel.”
Ask the question. “Sam, describe it to me the way you would to a table.” Listen to the whole answer without correcting mid-sentence.
Correct once, briefly, and only what matters. Usually that is the missing accompaniment or the missing allergen, not the adjectives.
Name the behaviour. “Tonight I want checkbacks inside two minutes of the first bite.”
Stop. Resist the fifth item, resist the story about last night, and resist the reminder that everyone has heard forty times.
That is under three minutes and it does more than the twelve-minute version, which is the part managers find hardest to believe until they run it for a fortnight.
Adapting it to different rooms
A high-volume casual restaurant needs the shortest version, and the question should be about speed-relevant facts: what is slow tonight, what has run out, what comes without a side.
Fine dining can afford a longer answer and should ask for precision, since guests are paying for detail. The question becomes “describe the third course” rather than “what is the special”.
A bar or cafe should ask about builds and specs rather than descriptions, because that is the knowledge that fails under volume.
Hotels and venues with a rotating floor need the written version more than anyone, because the person who missed Tuesday is working Thursday and there is no continuity without notes.
When people push back
Some floors resist being asked questions in front of colleagues, and the resistance is worth taking seriously rather than overriding.
Almost always it is fear of looking foolish, not objection to the format. That fades within about a week once people see that wrong answers are met with the answer rather than with a comment.
Start with the confident staff for the first few days. Once the pattern is established as normal and low-stakes, extend it to everyone.
Never start with the newest person. They are the ones who most need the practice and the least able to absorb an early public miss.
And if someone consistently dreads it, ask them privately rather than dropping the format. The problem is usually a specific gap they are embarrassed about, and it is fixable in five minutes one to one.
The measure that tells you it is working
Listen for the sentence quality rather than for correctness.
Early on you get menu copy read back, with the adjectives. Within a couple of weeks you get short spoken sentences with the accompaniment included. That shift is the whole return on the exercise, and you can hear it happening.
The second signal is that guests start being told what a dish comes with before they ask, which shows up as fewer sides added late and fewer disappointed faces when a plate lands.
The cost of not doing it
A five-minute briefing across six shifts is half an hour of paid floor time every week, multiplied by however many people are standing there. Most restaurants would scrutinise any other recurring half hour that produced nothing measurable.
The version described here costs exactly the same and produces a floor that can describe the menu, which shows up as fewer sent-back plates, fewer questions escalated to a manager, and better attachment on sides and drinks. None of that requires a budget line, a training day, or anyone staying late.
Frequently asked questions
What should a restaurant pre-shift meeting cover?
Three things: the one thing that changed today, one specific behaviour you want better tonight, and one question asked of one person and answered out loud. Cover numbers, procedural changes and individual feedback all belong elsewhere, because a briefing carrying nine items delivers roughly none of them.
How long should a pre-shift meeting be?
Five minutes, and shorter is usually better. The length is not the problem, though: a five-minute broadcast produces almost no retention, while a five-minute meeting containing one question asked and answered produces real recall. Cutting content is what makes the remaining content land.
How do I make staff actually remember the specials?
Ask one person to describe the special out loud rather than reading it to them. Everyone else answers silently while they wait, which is the mechanism that works. Rotate who gets asked so people prepare in advance, and ask about the same special again later in the week rather than only on the day it launches.
What is the best way to cover allergens in a pre-shift?
Make it a fixed thirty-second slot and rotate through dishes as questions rather than repeating a general warning. “Which of tonight’s specials contains dairy” teaches something; “remember to check allergens” stops being heard within a week. Pair it with a standing rule that “I will check” is always an acceptable answer.
What if half the floor starts later?
Run the same briefing twice at the two shift starts, identically. It costs the same five minutes and prevents a two-speed floor where the openers know the menu and everyone else reads the printed card. Write the three items somewhere late arrivals can read them, and ask one of them the question when they arrive.
