Guides
Twelve checks to run on brewery and taproom software before you buy
A vendor agnostic test of brewery and taproom software: twelve checks to run on a trial account, from exporting a day of sales to what you keep on exit.
What to take away
- No products are named or scored here. A software review written today is stale by the time you buy, and it cannot know what your operation needs to join together.
- What survives is the testtwelve checks you run yourself, on a trial account, using your own data.
- Judge systems on what they let you take out, not on what their dashboards put in front of you.
- Run the twelve checks on your current systems first. You may find the problem is a setting rather than a product.
- Decide in writing what you will do if the vendor disappears, before you sign anything.
Why this page names no products
A ranked list of software would be a snapshot of a market that changes constantly, filtered through whoever the writer happened to talk to. Worse, it would rank on features, and features are not where these systems succeed or fail for a brewery with a taproom.
What decides the outcome is whether four sets of records can be joined: sales, hours worked, payroll cost and beer produced. A product with an excellent feature list and no usable export will leave you retyping numbers into a spreadsheet every month, and you will stop doing it by March.
So this is a test you run rather than a verdict you accept. Twelve checks, an afternoon each, on a trial account.
The twelve checks
- Export a full day of transactionsas a plain file, with timestamps, without contacting support. If you cannot, stop here.
- Export hours workedat shift level, by person and by role, for a date range you choose.
- Record a comp and a sampleat the bar, attributed to a named authorizer, and time how long it takes. Anything over a few seconds will not survive a Friday.
- Name a product identicallyto how your production log names it, and confirm the two files can be joined on that name.
- Change a past recordand check whether the change is visible afterward, and to whom.
- Create a limited userwho cannot see payroll or void a sale, and confirm the limits actually hold.
- Turn on multi factor authenticationand check it applies to everyone, not only to the owner account.
- Handle a tipped transactionend to end and check what the export carries into payroll.
- Take the system offlineby disconnecting the internet and see what still works at the bar.
- Restore from a backupor export, and confirm you can read it without the vendor.
- Ask what happens on cancellationwhat you keep, in what format, and for how long.
- Price it at twice your current volume, including per terminal and per user charges.
Check nine matters more in this trade than in most. A taproom keeps serving when the connection drops, and a system that cannot take a payment or record a pour offline turns a network problem into a closed bar.
Twelve trial-account checks
- Export a full day of transactions
- Export hours worked by person and role
- Record a comp and a sample
- Match product names to production log
- Change a past record and trace it
- Create a limited user and verify limits
- Turn on multi factor authentication
- Handle a tipped transaction end to end
Check twelve catches the pricing model that is affordable now and painful later. Ask for the number in writing.
What to test with, and who should test
Use real data, not the demo set. The demo set is arranged to make the product look good, and your product names, your modifiers and your odd cases are what will actually break it.
Have the test run by the person who will use it most, not by the owner. A bartender will find in twenty minutes what a manager will not find in a week, because the failure modes are all about speed at the bar.
Then have somebody run the four way reconciliation described in the measures and the monthly reconciliation using only exports from the trial. If that cannot be done, the product does not fit, whatever else it does well.
Security is part of the evaluation
A small operation is a target precisely because nobody is watching, and the systems in question hold payment data, personal data and payroll.
The National Institute of Standards and Technology: NIST Small Business Quick-Start Guides break the basics into steps a business without dedicated staff can complete. The Cybersecurity and Infrastructure Security Agency: Cyber Guidance for Small Businesses covers the same ground from the attack side. That includes the account takeover and payment fraud patterns that hit small hospitality businesses.
Checks five, six, seven and ten above are the practical form of that advice. Add one process control that no software provides: remove access on the day somebody leaves. That is an offboarding step, so tie it to the checklist that comes out of hiring and training the crew.
The records the system has to support
Whatever you choose has to produce records you are obliged to keep, not only reports you enjoy reading.
Records the system must produce
Record type
- Income and expenses
- IRS
- Production and removal
- Your brewery
- Payroll and time
- Payroll and tax
- Licensee records
- State ABC board
- History retention
- Vendor plan
Who requires it
- Income and expenses
- Clearly shown, documents exist
- Production and removal
- Matches production log names
- Payroll and time
- Shift level, by role
- Licensee records
- Form and retention period
- History retention
- How long data is kept
What to confirm
- Income and expenses
- Production and removal
- Payroll and time
- Licensee records
- History retention
The Internal Revenue Service: What kind of records should I keep? guidance leaves the choice of system to you, provided it clearly shows income and expenses and the supporting documents exist. Alongside that sit your production and removal records, your payroll and time records, and whatever your state alcohol beverage control board requires of a licensee.
Ask each body what it expects and in what form, then check the product can produce it. Those obligations are part of what your license actually requires of you rather than an accounting preference.
Retention is the part people miss. Ask the vendor how long history is kept on your plan, because a system that quietly drops data older than a year cannot support a record you must hold for longer.
Scoring and deciding
Score each candidate out of two on every check: fails, works with effort, works cleanly. Then look only at the failures.
Deciding from the failures
Which checks failed?
Fails 1, 2 or 4 -> cannot get numbers out, reject
Fails 3 or 9 -> staff will resent it, weigh heavily
A product that fails checks one, two or four cannot be rescued by anything else, because you will never get the numbers out. A product that fails checks three or nine will be resented by the people who use it. Everything else is negotiable.
Do not buy for the operation you plan to become. Buy for the one you run now. Re run checks four and twelve before any change of scale; a growing operation is when a per terminal price or weak export starts to hurt.
If a bigger version of the business is on the table, that belongs with the routes to more volume rather than with the software decision.
Common questions
Do I need brewery specific software?
Not to begin with. A clean point of sale, a time clock that exports shift level hours, a payroll provider and a spreadsheet production log will run a small operation well. The signal to upgrade is the monthly reconciliation taking longer than an hour.
Is an all in one system better than joining several?
It depends entirely on its weakest module. All in one systems are usually strong in one area and adequate elsewhere, so ask which area is weak and whether that is the one you depend on.
How long should a trial be?
Long enough to include one busy night and one month end. A trial that never meets a rush has tested nothing.
What if the vendor will not let me export during a trial?
Treat that as the answer to check one and move on. Export access during a trial is the cheapest possible test of how the relationship will go.
Should the accountant choose it?
They should have a say on checks one, two and eight, since they will consume the output. The bar staff should decide on checks three and nine. Neither group should decide alone.
