Guides
A realistic take on brewery and taproom KPIs
Brewery and taproom KPIs on one page: six measures with written formulas and sources, what to leave off, and who should assemble the monthly report.
What to take away
- One page, six numbers, the same layout every month. A dashboard nobody reads is worse than a page somebody does.
- Every number needs a written formula and a named source. If two people would calculate it differently, it is not a measure yet.
- Put the comparison next to the number. A figure with no prior period beside it cannot change a decision.
- Split production and taproom in every line. Blending them produces an average that describes neither.
- Add a seventh line for the thing that went wrong this month, in words. Numbers do not carry causes.
The one page
Design it once and do not redesign it. The value of a monthly report comes from the same shape appearing every month, so that a change is visible without effort.
The One Page, Seven Lines
- Contribution per staffed taproom hour
- Labor as share of revenue, split
- Realized keg yield
- Depletion by beer, top and bottom
- Cash position, thirteen weeks forward
- Comps and samples, by authorizer
- What went wrong, two sentences
| Line | The number | Beside it |
|---|---|---|
| 1 | Contribution per staffed taproom hour | Same month last year, and last month |
| 2 | Labor as a share of revenue, taproom and production separately | Last three months |
| 3 | Realized keg yield | Rolling three month average |
| 4 | Depletion by beer, top and bottom three | Last month |
| 5 | Cash position and the next thirteen weeks | Last month's forecast against what happened |
| 6 | Comps and samples, by authorizer | Last three months |
| 7 | What went wrong, in two sentences | Whether it was on last month's page too |
Six numbers and one paragraph. The temptation is to add more, and adding more is how the page stops being read.
The formulas, written down
Write these on the back of the page so the definition travels with the number.
Contribution per staffed taproom hour. For each hour the room is open, take the revenue recorded in that hour, subtract the cost of the beer sold in it, and subtract the labor cost of the people on shift including payroll burden. Sum by day part and divide by hours. Source: point of sale, production cost records, payroll.
Labor as a share of revenue. Labor cost in the period over revenue in the period, calculated twice: once for taproom staff against taproom revenue, once for production staff against total revenue. Source: time clock and payroll against point of sale.
Realized keg yield. Pours sold from a keg, plus recorded comps and samples, divided by the theoretical pours in that keg. Theoretical pours are the keg's stated volume in gallons times one hundred and twenty eight fluid ounces, divided by your standard pour in ounces. Source: keg log and point of sale.
Depletion by beer. Volume sold per week by product, over four weeks. Source: point of sale, joined to the production log by a product name that matches exactly.
Cash position. Bank balance today, plus the thirteen week forward view. Source: bank and your own forward schedule.
Comps and samples. Recorded non sale pours, grouped by the person who authorized them. Source: point of sale.
The joining problem in line four is the one that breaks these reports in practice. If the taproom calls a beer one thing and the cellar calls it another, nothing will reconcile. Fix the naming once, at the source, and it stays fixed. The wider reconciliation this depends on is set out in the measures and the monthly reconciliation.
Why these six
They were chosen because each one can change an action next week.
What Each Line Decides
Line
- Taproom hour
- Keep, change or close hours
- Labor share
- Hiring and rostering
- Keg yield
- Recover beer already paid for
- Depletion
- What to brew, what to retire
- Cash position
- Prevent the cash crisis
- Comps and samples
- Training gap, inventory honesty
Action it changes
- Taproom hour
- Labor share
- Keg yield
- Depletion
- Cash position
- Comps and samples
Line one tells you which hours to keep, change or close. Line two tells you whether the schedule matches demand, which is a hiring and rostering question before it is a cost question, and it feeds directly into hiring and training the crew.
Line three is the closest thing this trade has to a free money detector. Yield loss is beer you already paid for in malt, labor and tank time, and small improvements move every pour you sell.
Line four decides what to brew next and what to retire. Line five is the one that prevents the crisis, since profitability and solvency are different questions and a growing operation can be short of cash precisely because it is growing.
Line six is not about suspicion. Unrecorded pours are noise that hides every other signal, and grouping them by authorizer usually reveals a training gap rather than anything else. Line six also feeds the quarterly stock review that keeps inventory counts honest.
What not to put on the page
- Anything nobody would act on. If the number moves and no decision changes, it does not belong.
- Revenue on its own. Revenue with no cost beside it has misled more owners than any other figure.
- Averages across production and taproom. They describe a business that does not exist.
- Week over week comparisons. In this trade those are mostly calendar and weather.
- Any benchmark from outside your operation. Your rent and wages are local, so a published percentage tells you about somebody else's building.
Reading it honestly
Three habits keep the page useful.
Reading It Honestly
- Compare like with like, note holidays
- Look at the trend before the level
- Ask what would have to be true
- Check the cheapest cause first
Compare like with like. A month containing a holiday weekend or a festival is not comparable to one without, and the honest response is to note it in line seven rather than to quietly explain it away.
Look at the trend before the level. One month is a data point. Three months in the same direction is a finding.
Ask what would have to be true. If line one fell, list the possible causes, then check the one that is cheapest to verify first. Usually that is the schedule rather than anything more interesting.
Where the numbers come from and who owns them
Name an owner for each line. Not a committee. The owner is responsible for the number being produced on the same day each month with the same method.
Who Owns Each Number
Name one owner per line, not a committee
same day, same method, every month
number drifts and stops being comparable
Have the report assembled by somebody other than the person who enters most of the underlying data. That is not about distrust, it is that a second pair of eyes catches the transcription error the first pair made. It also means the report survives a holiday.
The records feeding these numbers are the same ones you are obliged to keep. The Internal Revenue Service: What kind of records should I keep? guidance leaves the system to you as long as it clearly shows income and expenses with supporting documents behind it, and the retention obligations your license carries are part of what your license actually requires of you. Build the report out of records you keep anyway rather than creating a parallel set.
Protecting the numbers
The systems producing these figures hold payment data and payroll data, and small operations are targeted because nobody is watching.
The National Institute of Standards and Technology: NIST Small Business Quick-Start Guides set out the basics for a business without dedicated staff, and the Cybersecurity and Infrastructure Security Agency: Cyber Guidance for Small Businesses covers the fraud patterns that reach small hospitality businesses. The parts that matter here are individual logins, multi factor authentication on anything touching money, and visibility of who changed a past record. A number that can be silently edited is not a control.
When to change the page
Rarely, and deliberately. Change it when the business changes shape: a second room, a packaging line, a distribution relationship, or any of the routes to more volume. At that point add the one line the new activity needs and retire one that has stopped changing decisions.
Resist adding a line because a system offers it. The constraint on this page is attention, not data.
Common questions
Is monthly often enough?
For these six, yes. Line five deserves a weekly glance, and line one is worth looking at by day part whenever you are considering a schedule change.
What if I do not have the data for one of these?
Then that is the first thing to fix, and it is usually a naming or an export problem rather than a missing system. Start the line with a rough figure and improve the method rather than waiting for perfect data.
Should staff see the report?
The parts that concern their work, yes, and it generally helps. People schedule and pour differently when they understand what a slow hour costs and what a yield figure means.
How many months before it is useful?
Three to see a trend, twelve to understand your seasonality. Keep every page, because next year's plan is built from this year's file.
What if the numbers contradict what I feel about the business?
Check the formula and the source first, then trust the numbers. The most common cause of the contradiction is a definition drifting, which is why the formulas belong on the back of the page.
