Bar Revenue Playbook: Find the money being left on the table

The Revenue Walking Out the Door

You watch your revenue closely. You know what last Saturday did, and how it compares to last week and last year. The figure almost nobody works out is the gap between what you took in and what the building was set up to take in.

That gap sits in a few specific places: servers who take orders rather than sell, a menu priced on instinct, an event room that nobody is responsible for, and a schedule built from memory. None of it appears as a line on your profit and loss statement. It appears as a revenue number that never quite reaches where it should.

A bar with no revenue systems in place typically leaves 8 to 15 percent of what it could make. On 750,000 a year that is 60,000 to 110,000, from the building you already have and the guests already sitting at your tables. You can capture it, and the rest of this page explains how.

The Four Gaps Running at Once

Sit down with a busy bar whose revenue is not growing and you find the same four gaps running together. Each one loses money on its own. All four at once is why the room can be full and the number still will not move.

Check Average Is the Fastest Thing You Can Move

Average check is what each guest spends per visit. Raising it needs no new customers and no bigger room. It means the people already at your tables spending three to five dollars more each. Put in how many guests you serve in a day and see what that is worth over a year.

What 90 Days Looks Like

This is a worked example, not a promise. Take a full-service bar and kitchen doing about 1.2 million a year. The owner believes her average check is around 30 dollars. Pulling three weeks of sales by server shows 31.40, with three servers under 26 and two over 38. The menu has never been checked for margin. Labor is at 26 percent against a 21 percent target. The event room books four a month.

The first measurement is always the same: the gap is bigger than anyone expected. That is not a problem with the bar. It is the first accurate picture it has had of itself.

Week two: a short briefing runs before every service with a check average target and two featured items. The first pass over the menu moves three high-margin items into better positions and lists the ones priced below what they cost.

Week six: the check average is climbing. One server comes off a dead Thursday. The event room has a rate card and a rule that every inquiry is answered within two hours.

Day 90: check average up 3.80. Floor labor down to 22 percent. Nine events booked in the month against four before. At that rate over a full year, that is about 147,000 captured. It is the same room, the same staff and no new customers. The only thing that changed is that these systems now run every week.

Recovery Has a Timeline

The first 30 days are about getting your starting numbers, not results. Those first numbers are your baseline, which is what everything later gets measured against. You work out the average check for each server for the first time and find a spread you did not know existed. You go through the menu for margin and see which items are carrying the others. You split labor by department and find the floor a few points over.

Days 30 to 60 are where the first numbers move. The briefing is running, the average check usually starts climbing within three or four weeks, the schedule gets built from a forecast instead of from memory, and event inquiries get answered the same day.

Days 60 to 90 are where the systems start to add up. The spread between your servers narrows because of the coaching, the events you booked earlier start arriving, and the effect shows up in the Monday numbers.

Most bars stop somewhere between day 20 and day 45, when the briefing gets skipped on a busy night and the Monday review slides to Tuesday. Whatever you use to run this, you need something that tells you a system has stopped running before the month ends.

How Much Revenue Are You Capturing?

Ten questions, yes or no, with no partial credit. If the honest answer is somewhere in between, that is a No. Either the system is running this week or it is not.

Most operators know their revenue figure exactly. Ask for the average spend per guest last Tuesday and the room goes quiet. Ask what floor labor ran as a percentage of revenue last week and the subject changes. That gap is what this diagnostic measures. The monthly figure beside each No is a common cost for that gap, not a measurement of your bar.

Every Gap: Where It Sits, What Closes It

Before the systems themselves, here is the whole picture on one screen. For each gap: the work that captures it, the number that shows it to you, and the system on this page that closes it.

The paper these systems need is on this page too. The server standards and scripts, the pre-shift briefing form, and the table visit audit each sit inside the system that uses them, and you can read or print any of them for free.

The Numbers to Run Against

These are your reference points. Find the measure you are looking at, know what it should be, and know the point where it turns into a problem. Put your own figure in below and it will tell you where you sit.

Menu and Pricing

Labor Cost and Productivity

Check Average and Selling

Seven Systems, One Revenue Engine

These are not separate fixes. Sorting your menu by margin without also fixing the prices leaves you with the right items in the right positions at the wrong price. Building a schedule without checking what each hour brings in gives you a schedule that feels right but has no connection to the revenue it has to support. Teaching the floor to sell without measuring the average check gives you activity with no way of knowing whether it worked.

The systems are in this order on purpose. Menu engineering comes first because every pricing decision, every suggestion at a table and every event menu is built on knowing which items make you money and which ones you are subsidising. Each system produces the information the next one needs.

Menu Engineering

A bar in Atlanta, 72 seats, the same menu for three years. Salmon was her third-highest seller. She was proud of it, put it in the feature box, and trained her servers to push it. Running the numbers showed the salmon made her 7.20 a plate. Her ribeye, which she never pushed because it felt expensive to suggest, made her 19.40. Same covers, same kitchen. She had been training her team to sell her worst-margin item, and losing 12 dollars every time they succeeded.

Food cost as a percentage tells you whether an item is priced sensibly against what it costs. The money that reaches your bank account is the margin in dollars: the menu price minus what the ingredients cost you. Most operators promote the items that sell the most rather than the items that make the most, and those are rarely the same thing. Filling your best menu positions with anything other than your best-margin items is money walking out the door.

Every item falls into one of four groups. Items that make good money and sell well are your stars: feature them and name them in the briefing. Items that sell well but make little need either a price rise or a cheaper recipe. Items that make good money but sell little usually have a visibility problem, so move them and rewrite the description. Items that do neither get reworked or cut. Putting a box around an item or giving it a proper description moves how often it gets chosen by 20 to 30 percent, and that costs the kitchen nothing.

Run It Without Software

A spreadsheet with four columns will do this. The arithmetic is simple. What takes the time is costing the recipes honestly.

1. List every menu item with its selling price. Start with food and drink separately, because their costs work differently, and mixing them together makes the ranking meaningless.
2. Work out what each item costs you to make. Add up every ingredient at what you actually pay, including the parts you throw away. A protein that is only 70 percent usable after trimming costs you the purchase price divided by 0.70, not the purchase price.
3. Work out the margin in dollars. Selling price minus what it costs you. The trade name for this figure is contribution margin, and it is worth knowing because you will hear it used. It is the column most menus have never had, and it is the one that matters: a percentage tells you how well an item is priced, but the dollars are what actually reach your bank account.
4. Add how many of each you sold last month. Your POS will give you this. Product mix, item sales, or sales by item are the usual names for the report.
5. Work out two averages. The average margin across all your items, and the average number sold. These two figures are the lines that split the four groups.
6. Sort every item into its group. Above both averages, above one, or below both. Write the group next to each item.
7. Act on the group, not on your feelings about the dish. Good margin and good volume go in the best positions. Good volume and poor margin get a price rise or a cheaper recipe. Good margin and poor volume get moved and rewritten. Poor and poor get one 90-day test in a better position before you cut them.
8. Redo it every quarter, and re-cost any item whose ingredients moved more than 5 percent. An out-of-date cost puts an item in the wrong group, which leaves you worse off than not sorting them at all.

Every Quarter

• Re-cost every item whose ingredients moved more than 5 percent since the last review.
• Sort the whole menu by margin in dollars, not by food cost percentage.
• Check that every high-margin item is in a position people actually read.
• List the high-volume, low-margin items for a price change or a recipe change.
• Put the poor performers on a 90-day test in a better position with a better description, then decide.
• Give every high-margin item a description that names its main ingredient and a pairing.

Pricing

A bar owner in Denver, full-service, good reviews, steady regulars. He had not changed a menu price in 26 months. His chicken entree was still 18 dollars while what he paid for the protein had gone up 21 percent and his kitchen labor 14 percent. The food cost on that plate was now 44 percent. He was serving it to 80 people a week and losing ground on every one, assuming the menu was fine because nobody had complained.

Most operators do not have a pricing approach, they have a pricing history. Prices were set when the place opened, nudged when costs got painful, and otherwise left alone because raising them feels risky. Pricing that only reacts will always lag behind your costs. What fixes it is a review on a fixed calendar, which catches an increase in April rather than finding it on the May statement.

Do not set your prices by looking at the bar down the street. Their supplier discount, their rent and their prep cook are not yours. Their 16 dollar burger might run a 28 percent food cost while yours at the same price runs 34. Price against your own cost floor, which is what the ingredients cost you divided by the food cost percentage you are aiming for.

Guest resistance is also wildly overestimated. A one to two dollar increase on a mid-menu item, made during a normal reprint rather than announced, almost never gets commented on. What does get noticed is a large jump after two years of nothing, which is exactly what happens when small corrections are avoided.

Run It Without Software

1. Work out the price floor for every item. Take what the ingredients cost you and divide it by your target food cost percentage written as a decimal, so 30 percent becomes 0.30. An item costing 6 dollars with a 30 percent target has a floor of 20 dollars. Anything priced below its floor is losing you money on every plate.
2. List everything sitting below its floor, and everything within a dollar of it. That list is your work, in order of how many you sell.
3. Work out the break-even before you raise anything. A price rise pays for itself even if you lose some volume, and the calculator below tells you how much you could afford to lose.
4. Move prices during a normal reprint, not as an announcement, and change several items at once rather than one at a time.
5. Price drinks, wine, draft beer and food separately. Their costs work differently, and one target across all of them will price one too high and the other too low.
6. Put a date on the calendar every quarter, and review immediately whenever a supplier raises a high-volume item by more than 8 percent.
7. Finish the review with a written list and a date the new menu prints. Deciding that you will get to the salmon at some point is not a review.

Every Quarter

• Work out the price floor for every item whose ingredient costs have moved.
• List everything priced below its floor, in order of how many you sell.
• Work out the break-even on each proposed rise before you decide.
• Price drinks, wine, draft and food on their own targets.
• Review immediately if a supplier raises a high-volume item more than 8 percent.
• Close the review with a written list and a date the new menu prints.

Labor Cost and Scheduling

A full-service concept in Chicago, 110 seats. On a Tuesday in February there were 22 servers on for a shift that did 4,200 dollars in food and drink. That is 191 dollars a server, with floor labor at 58 percent of revenue. The manager who built it did what he always did on Tuesdays, the same way, for two and a half years. Nobody had ever told him to look at the expected revenue before he wrote the names down.

Labor is the largest expense you can actually control. Your rent does not move with your revenue; your labor should, and in most bars it does not, because the schedule gets built the same way every week. The fix is not cutting staff or wages. It is building the schedule from a revenue number instead of from memory.

A single combined labor percentage hides the problem. A total of 32 percent can be a tight bar carrying a kitchen and a floor that are both five points over, and you will not see that until you split them.

Overtime is a scheduling failure rather than a staffing solution. A handful of people at five premium hours a week runs well into five figures over a year, and you are paying a premium on hours you were always going to have. Slow shifts are the other quiet loss: Monday and Tuesday often do half the weekend's revenue on three-quarters of the people, because the extra body feels safer than sending someone home.

Run It Without Software

1. Forecast the revenue for each day before you build anything. The same day last week and the same day last year, adjusted for anything you know about, is accurate enough to work from.
2. Turn that into a labor budget in dollars. Forecast revenue multiplied by your target labor percentage for that department.
3. Turn the dollars into hours. Labor budget divided by your average hourly rate for that department, including what the employer pays on top of wages. That hours figure is what you are allowed to schedule.
4. Build the schedule to the hours, writing names only after you know the number. If you go more than 5 percent over on any shift, settle it before you post the schedule rather than after.
5. Keep bar, kitchen and floor as three separate numbers all the way through. A combined figure cannot tell you which one is over.
6. Every Monday, compare what you actually spent to the target, by department, and flag anything more than two points over.
7. Decide whether a miss was a scheduling error or a revenue miss, and write down one action before you close the review. They need different fixes and treating them the same is how the same miss repeats.
8. Treat overtime as a schedule to fix. Move the hours to someone under 40 rather than accepting the premium.

Every Week

• Forecast revenue by day, and work out the labor budget in hours, before building anything.
• Build to the hours. Settle any shift more than 5 percent over before you post it.
• Monday: compare actual labor to target by department, not as one figure.
• Flag anything more than two points over target.
• Decide whether each miss was a scheduling error or a revenue miss, and write down one action.
• Move overtime hours to someone under 40 rather than accepting the premium.

Revenue Per Labor Hour

Two bars in the same city, similar concepts, both around 900,000 a year, both running 28 percent labor. The first one brings in 68 dollars of revenue for every labor hour on a Saturday and 41 on a Tuesday. The second brings in 42 on a Saturday and 28 on a Tuesday. Identical labor percentages. The first gets 60 percent more revenue out of every labor dollar, and the percentage hides that completely.

Revenue per labor hour is what you get back for each hour you schedule: the shift's revenue divided by the hours it used. Labor percentage answers whether you spent the right share of your revenue. Revenue per hour answers whether that spend actually produced anything. A Saturday running 38 against a 55 target is a 17 dollar gap every hour, and across a season of Saturdays that is real revenue the staff you are already paying for are not capturing.

A low figure has three possible causes and each needs a different fix. Too many people scheduled means rebuilding from the forecast. An average check that is too low means the briefing and what gets offered at the table. Revenue that simply came in under forecast is usually a one-off and needs nothing. Getting the diagnosis right matters, because cutting the schedule to fix a selling problem just makes the service worse and the selling no better.

Run It Without Software

1. For each shift, take the revenue and divide it by the hours worked. Use scheduled hours if that is all you have, and actual clocked hours if you have them, but be consistent about which.
2. Do it by shift, not by week. A weekly figure averages a strong Saturday against a weak Tuesday and hides both.
3. Set your target from your own best four weeks, not from a benchmark in a guide written about a different concept. Take your best four weeks, work out the figure for each shift, and add 10 to 15 percent.
4. Flag any shift more than 10 percent below its target, and look at the four-week trend for it rather than the single week.
5. Diagnose before you act. Compare the covers and the average check for that shift against a good one. If the covers were there and the check was low, it is a selling problem. If the covers were not there, it is a scheduling problem.
6. Build the next schedule from the revenue the shift is set up for, working backwards to a headcount rather than starting from a headcount habit.

Every Monday

• Work out revenue per labor hour for every shift last week.
• Flag any shift more than 10 percent below its target.
• Look at the four-week trend for each flagged shift, not the single week.
• Work out whether it was a scheduling problem or a selling problem before acting.
• Build next week from the revenue each shift is set up for.
• Review your targets against your own best four weeks each quarter.

Check Average and Upselling

A full-service bar in Portland, 85 seats. The owner thought her average check was around 30 dollars, maybe a little more. Three weeks of sales pulled by server showed her highest at 41.20 and her lowest at 23.80. Same menu, same guests, same room. A 17.40 spread between her best and worst server that she had never seen, because she had only ever looked at total revenue rather than at what each person was generating per guest.

Average check is the revenue each server brings in per guest, and it is the number that tells you whether your floor is selling or taking orders. Two servers with the same section produce very different revenue if one suggests a starter and a dessert and the other asks whether there is anything else.

The wording matters more than most people expect. A server who says there is a good Old Fashioned tonight is making a suggestion. A server who asks whether you want a drink is making an offer. Specific suggestions get taken up at two to three times the rate of open questions, and that difference costs nothing to train.

A three dollar lift across the floor needs no new customers and no bigger room. What it needs is a briefing before each shift and a number to aim at. The briefing is the highest-payoff five minutes in the building, and it is operational rather than motivational: the featured items, the target for the shift, the order things get offered in, and one pairing everybody can name.

Run It Without Software

1. Get sales and covers by server out of your POS. Almost every system has this report and in most bars nobody has ever run it. Server sales, server performance or employee sales are the usual names.
2. Divide each server's sales by their covers. That is their average check. Sales alone will not do: a server with more tables always shows more sales.
3. Work out the team average and the percentage each server sits above or below it.
4. Do it weekly, on the same day, and keep the sheet. One week is a reading; four weeks is a pattern you can coach from.
5. Run a briefing before every service. Five minutes: two featured items and why, the check average target for the shift, the order things get offered in, and one drink or wine pairing everybody can name.
6. Write down the order things get offered in and hand it to every server. If it is not written down it is not a standard, it is a preference.
7. Check a live table twice a week at unpredictable times. Sit where you can hear a greeting and an order being taken, and write down what you heard against the sequence.
8. Flag anyone more than 15 percent below the team and coach them this week rather than next.

Every Week

• Run a briefing before every service: featured items, a target, the sequence, one pairing.
• Work out the average check for every server and compare it to the team.
• Flag anyone more than 15 percent below the team average.
• Anyone trending down two weeks running gets a conversation this week.
• Sit in on two live tables at unpredictable times and write down what you heard.
• Update which items are featured, so the briefing names your best-margin dishes.

The Paper This System Needs

Server Performance

A general manager at a 110-seat place spent three years watching his best server and his second-best from across the room. Both were favorites with the regulars, both professional. He asked which one to promote to floor lead. Asked what their average checks had been over the last 60 days, he had never pulled the number. Working it out together showed his best server at 29 dollars a guest and his second-best at 41. He had been about to promote the wrong person, because he was managing on impression rather than on what each of them actually brought in.

The floor generates 60 to 70 percent of revenue in most full-service places and gets the least scrutiny of anything in the building. You track what your product costs to the penny and let the floor run on personality. Two reliable, well-liked servers sitting a few dollars below the team average can quietly cost tens of thousands a year, not because they are bad at the job, but because nobody ever told them the number.

Coach from the number rather than from a judgement. Saying that their tables are not selling enough is an opinion, and an opinion is something to argue about. Saying that their average check over four weeks was 24.80 against a team average of 33.40 is a fact you can both work from. Most servers below the average are not lazy. They are uncomfortable at one specific point, usually offering dessert, and nobody has ever helped them get comfortable with it.

A written standard, signed when someone is hired, is what makes every later conversation fair and enforceable. And naming the top performer each week in the briefing tells the whole floor that the number is being looked at.

Run It Without Software

1. Keep the weekly sheet from System 5 with each server's average check and how far they sit from the team. Four weeks of it is what you coach from.
2. Sort by the gap, not by total sales. Total sales rewards whoever had the most tables, which is a rota decision rather than a performance one.
3. Work out what each gap costs over a year before the conversation. It changes how seriously both of you take it.
4. Open the conversation with their number and the team number, in that order, and nothing else. No adjectives.
5. Find the touch point they skip. Ask them to walk you through a table from greeting to check. The one they rush past is almost always the one costing them the gap.
6. Agree one specific change and a date to look again, then write both down.
7. Check the four-week trend two weeks later. If it is moving, say so out loud. If it is not, have the second conversation in writing.
8. Name the top performer each week in the briefing. It costs nothing and it tells everyone the number is seen.

Every Week

• Work out the average check for every server and mark anyone more than 15 percent below the team average.
• Name the previous week's top performer in the next briefing.
• Book coaching for below-average servers this week, not next, and open with the number.
• Hand out the written standard and keep a signed copy for everyone on the floor.
• Check the four-week trend two weeks after each conversation.
• If it has not moved, have the second conversation in writing.

Putting It In Place and Keeping It Running

A bar owner in Seattle reads every book about running a better restaurant. She highlights them, she folds the corners, she starts about twice a year and gets through week one with real energy. Then a Saturday goes wrong, a manager calls in sick, and the spreadsheet she opened on Tuesday sits untouched until she finds it three months later. The information was never the problem. What was missing was a plan in a set order, with a name against each task, a date on it, and one number to check that it is working. Motivation runs out after about ten days. A written routine does not, because it does not depend on how anyone feels that week.

Most attempts stop at around day 45 rather than in week one. A briefing gets skipped on a busy night, the Monday review slides to Tuesday, the audit sheet sits in a drawer. Each one feels like a single exception at the time and none of them are.

The order matters as much as the content. Menu engineering comes before pricing, labor cost before revenue per hour, and the average check before the written selling standard, because each one produces the information the next one needs.

The First Four Weeks

Week 3 is uncomfortable the first time you hand experienced servers a written standard. Do it anyway, and do it directly rather than apologetically. This is how the floor runs now.

Work the Rollout

Put a name and a day against each of these before you start. A task with nobody's name on it only gets done when somebody feels like doing it, and that is exactly the failure this system exists to prevent.

• Week 1: Cost your menu and sort it by margin in dollars.
• Week 1: Pull sales and covers by server and work out everyone's average check.
• Week 1: Set labor targets by department and a revenue-per-hour target by shift.
• Week 2: Work out the price floor for every item and list what sits below it.
• Week 2: Build next week's schedule from a revenue forecast instead of from last week.
• Week 3: Hand out the server standard and collect a signature from everyone.
• Week 3: Run the first pre-shift briefing, and keep the sheet.
• Week 3: Hold the first coaching conversation, opening with the number.
• Week 4: Run every system in the same week and hold the Monday review.
• Day 45: Check which system slipped first. It is always one of them. Restart it.

The Revenue Is Already in Your Building

A bar owner in Asheville, 78 seats, three months in. Check average up 3.80 from where she started, at 110 covers a night. Nine private events in October against four the month before. Floor labor at 22 percent against the 29 percent her habit-built schedule had been running. She said she could not believe none of it was complicated. It was not. She had simply never had a system that made her look at it every week.

What separates bars that capture their revenue from bars that leave it behind is not location, concept or talent. It is measuring, and having a system instead of a feeling. A briefing before every service. A schedule built from a revenue number. A rate card that goes out within two hours. None of it is complicated and all of it needs a system.

The gap between what you are making and what you should be making is not a problem with your market. It is a problem with your systems. Start building them tonight.