A restaurant dashboard you look at daily

A restaurant dashboard is a single screen that shows, at a glance, how the venue is trading and what needs attention today. It is not a report, and it is not a place to explore. It is the thing a manager opens with the first coffee of the morning, reads in a minute, and closes knowing whether yesterday went to plan and whether today needs anything unusual. If it takes longer than that, or if nobody opens it, it is not doing its job.

Most dashboards fail in the same way: they grow. Every manager adds a number they care about, every new feature adds a chart, and within a year the screen holds thirty tiles that nobody reads. The useful dashboards are the ones that stay small, and staying small takes discipline: a clear purpose, a fixed number of tiles, and a rule that anything new must replace something old.

This guide sets out a daily dashboard of six tiles, what each one is for, how to read the whole screen in a minute, and why the seventh tile should be refused. It covers how the dashboard relates to the weekly KPIs and the reports behind them, how a group can use one consolidated view across venues, and where scheduled reports fit in.

A daily dashboard answers three questions. How did yesterday go? What does today look like? Is anything wrong that needs action before service? Everything on the screen should help answer one of those three, quickly.

It does not answer why. When a tile shows something unexpected, the manager goes to the report behind it, or raises it at the weekly review. The dashboard is a smoke alarm, not a fire investigation. Keeping that distinction clear is what keeps the screen small.

Six is a practical limit. It is enough to cover sales, guests, spend, bookings, reliability and one operational signal. It is few enough to read in a minute and to remember: after a few weeks, a manager knows the six tiles so well that they notice a change without reading the numbers closely.

The exact number matters less than the fact that there is one. A dashboard with a fixed limit forces choices; a dashboard without one grows until it fails. Six is a good default. A small venue might use four; very few need more than eight.

The word comes from the car. A driver's dashboard shows speed, fuel and a few warning lights, and nothing else, because a driver needs to glance at it without taking their eyes off the road for long. Software borrowed the idea: a dashboard in computing is a screen that brings the most important measures together so they can be read at a glance.

The useful part of the comparison is what a car leaves out. There is no chart of last month's fuel use on the dashboard, and no list of every journey. Those belong elsewhere. A restaurant dashboard works best when it follows the same rule.

If the venue has no dashboard yet, start on paper. Write down the six numbers you would want to know every morning, and for each one, where it comes from and what you would compare it with. Then try reading them every morning for two weeks, even if someone has to put them together by hand.

After two weeks, you will know which numbers led to a decision and which were simply read. Keep the useful ones, replace the rest, and only then build the screen. A dashboard designed from two weeks of real mornings is far more useful than one designed from a list of everything the system can show.

Net sales for yesterday, against the same day last week and the same day last year. The comparison with the same weekday matters: Tuesday against Tuesday, not Tuesday against Monday. Show the difference as a percentage, with a simple colour for whether it is within the usual range.

This is the headline, and it is where most managers look first. On its own it says what happened, not why. The next two tiles help explain it.

The number of guests served yesterday, against the same day last week and last year, ideally split into lunch and dinner. Covers show whether more or fewer people came. Read with tile one, they show whether a change in sales came from the number of guests or from what they spent.

If sales fell and covers held steady, spend fell. If both fell, fewer guests came. Each points to a different question and a different owner.

Sales divided by covers for yesterday, by service. This completes the explanation that tiles one and two begin. A fall in spend at dinner, with covers steady, points to the menu, the drinks or the team's selling; a rise after a menu change shows the change working.

Keep it split by service. A single figure for the whole day blends lunch and dinner and moves when the mix changes, even if nothing else did. Our guide to average spend per head, split properly covers why.

Covers already booked for today and each of the next six days, against the same point last week. This is the only tile that looks forward, and it is often the most useful. A busy Thursday spotted on Monday can still be staffed and ordered for; a quiet Saturday spotted on Monday can still be filled.

Show it as a simple row of seven days, with each day's booked covers and a colour for whether it is ahead of or behind last week. Our guide to a covers forecast your kitchen will believe covers how to turn the booked figure into a forecast.

The number of booked covers who did not arrive yesterday, and the number who cancelled late, with the rate against recent days. This is the reliability tile: it shows whether the book can be trusted, and it gives early warning of a rise that needs attention.

A single bad night is not a trend. Three in a week on the same service is. Our guide to your no-show rate covers how to measure it fairly.

The sixth tile is for the one operational measure that matters most in this venue. For a busy kitchen, it may be ticket times yesterday: the median time from order to pass, against the target. For a neighbourhood venue built on regulars, it may be the share of covers from returning guests. For a venue that relies on feedback, it may be yesterday's average rating.

Choose one and keep it. The temptation is to show all three, which is how dashboards grow.

Read the tiles in order. Sales: up or down? Covers: up or down? Spend per head: up or down? Together, these three explain yesterday in a sentence. Then the booked covers: anything unusual in the next week? Then no-shows: was yesterday normal? Then the operational signal: was service as it should be?

With practice, this takes under a minute. Anything unusual becomes a question for the right person: the chef, the floor manager, the bar manager. Most days, there is nothing to do, and that is the point.

Sooner or later, someone will ask for a seventh tile: labour hours, food waste, website visits, the weather, the number of reviews this week. Each request is reasonable on its own. Together, they turn the dashboard into a report that nobody reads.

The rule is simple: a new tile must replace an old one. If the new number is more important than one of the six, swap them. If it is not, it belongs in a report or the weekly review. Most requests, examined this way, turn out to be weekly questions rather than daily ones.

Many important numbers do not belong on a daily dashboard because they are not useful daily. Food cost percentage, labour cost percentage and prime cost are best read weekly, when the noise of a single day has been smoothed out. Guest retention and lifetime value move slowly and are best read monthly.

These numbers belong on the weekly KPI sheet and in the monthly accounts. Putting them on the daily screen adds noise and makes the important daily numbers harder to see.

The daily dashboard and the weekly restaurant KPIs work together. The dashboard catches daily surprises; the KPI sheet shows weekly trends and the costs that matter. Some numbers appear on both, such as covers and spend per head, but they are read differently: daily for surprises, weekly for trends.

Our guide to restaurant KPIs worth tracking every week sets out the weekly list, and our guide to a weekly review that takes thirty minutes covers the meeting where it is read.

Behind every tile is a report. The sales tile sits on top of the sales report; the covers and no-show tiles on top of the booking reports; the operational tile on top of the kitchen or guest reports. When a tile shows something unexpected, the report behind it explains why.

Good restaurant reporting keeps those reports few and consistent. Our guide to restaurant reporting: six reports, no more sets out the reports that support a dashboard like this one.

A number on its own means little. Each tile needs a comparison: the same day last week, the same day last year, or recent days for rates. Choose the comparison that removes the most noise. For sales and covers, the same weekday last week is usually best for daily reading; last year adds the seasonal view.

Show the comparison as simply as possible: a percentage, an arrow, a colour. A manager should see the direction at a glance without reading the numbers.

Give each tile a range: a band where the number is normal, and bands above and below where it deserves a question. Colour the tile by where today's number falls. Green means nothing to do. Amber means look. Red means act.

Keep the ranges wide enough that most days are green. A dashboard that is amber every day trains people to ignore the colour. A dashboard that is usually green, with an occasional amber, gets attention when it matters.

Some dashboards also show what is happening right now: guests seated, guests waiting, tables in service, open orders, today's covers so far. This live view is useful during service, especially for a manager who is not on the floor, but it is a different tool from the morning glance.

Keep the two apart in your head. The morning dashboard is about yesterday and the week ahead; the live view is about the next hour. Both are useful; neither should be cluttered with the other's numbers.

Many managers read the dashboard on a phone, on the way in or between tasks. Design for that: tiles large enough to read on a small screen, numbers rather than detailed charts, and the most important tiles first. If the dashboard only works on a large screen, it will be read less often.

The same discipline applies. Six tiles fit on a phone; thirty do not.

In a group, the owner or operations lead needs a consolidated view across venues as well as each venue's own dashboard. Consolidated reporting shows the same tiles for the whole group, with the option to filter to a single venue or compare them side by side.

Keep the same tiles at every level, so a venue manager and the operations lead are looking at the same measures. Where venues trade in different currencies, the group figures need converting to a single currency at a stated rate, or the totals are meaningless.

A group view adds things a single venue does not need: which venues are trading above or below their usual range today, which have unusual bookings ahead, and which have setup or service problems. It should highlight the exceptions rather than show every venue's every number.

An operations lead with eight venues does not have time to read eight dashboards every morning. A group view that says "two venues need a look today" is worth more than one that shows everything.

Some venues prefer scheduled reports: a summary emailed each morning at a fixed time, so the numbers arrive without anyone opening a screen. Others prefer to open a dashboard. Both work, as long as the habit is daily and the content stays small.

The danger with scheduled reports is the same as with dashboards: they grow, and an emailed report with thirty numbers is easy to ignore. Whichever form you choose, apply the same rule. Six numbers, one glance, and a new one only replaces an old one.

The general manager or owner looks at the dashboard every morning. The head chef might look at the covers booked ahead and the operational tile. The front of house manager looks at covers, no-shows and bookings. In a group, the operations lead looks at the consolidated view.

Each person reads the tiles that matter to them, but they all see the same screen. That shared view helps: when the chef and the floor manager both know Thursday is busy, they plan together.

When a tile shows something unusual, act in the right place. A sales fall with steady covers is a question for the floor manager and the bar. A covers fall is a question about bookings and marketing. A rise in no-shows is a question about reminders and confirmations. A ticket time problem is a question for the kitchen.

Most changes need a short conversation the same day, and a note for the weekly review. Very few need a meeting.

Once or twice a year, review the dashboard. Has each tile led to a decision in the past few months? Is there a tile nobody looks at? Has the business changed, with a new service, a new venue or a new problem, so that a different tile would serve better?

Change the tiles deliberately, and tell everyone who uses the dashboard. A dashboard that changes often loses the familiarity that makes it fast to read.

Here is an illustrative morning for a manager reading a six-tile dashboard. The figures are examples.

Actions: the manager asks the chef about the dinner peak and suggests an extra cook on Thursday; a short email to regulars for Saturday is scheduled. The whole reading took about a minute.

In servearound, each venue's dashboard shows revenue, covers, spend per cover and no-shows against the previous period, a revenue and covers trend, the busiest hours and the split between booked and walk-in guests. A "Right now" panel shows covers and bookings today, guests seated and waiting, tables in service and open orders. Every day and hour is on the venue's local clock, and it works in a browser, on iPhone and iPad, and on Android.

For a group, the network dashboard brings the same measures together for all venues or one, for a quick range or a custom one, with revenue and covers trends, an activity feed and figures such as busiest table, returning guests, ticket time and top seller. servearound does not email scheduled reports: the dashboards are there whenever you open them. See Analytics or get in touch.

Six is a good default. A small venue might use four; very few need more than eight.

Yesterday's sales, covers and spend per head, covers booked for the week ahead, yesterday's no-shows, and one operational signal such as ticket times.

No. Food cost, labour cost and prime cost are best read weekly, when daily noise has been smoothed out.

The general manager or owner every morning, with the chef and front of house manager reading the tiles that matter to them.

Either works if the habit is daily and the content is small. The rule about six numbers applies to both.

With a consolidated view across venues that highlights exceptions, and the same tiles at venue level.

It must replace an existing one. If it is not more important than any of the six, it belongs in a report or the weekly review.

Rarely. Review them once or twice a year, and change them only when the business has changed.

It should show ranges, through colour, rather than single targets. A range tells the reader at a glance whether a number is normal, worth a look or worth acting on.

A restaurant dashboard you look at daily is a single screen of six tiles: yesterday's sales, covers and spend per head, covers booked for the week ahead, yesterday's no-shows, and one operational signal. Each tile has a comparison and a range, and the whole screen is read in a minute.

Refuse the seventh tile: anything new must replace something old. Keep weekly measures on the KPI sheet and the detail in the reports behind each tile, give a group a consolidated view that highlights exceptions, and whether you open a screen or receive scheduled reports, keep the habit daily and the content small.