On this page
Key takeaways
- Onboarding gets faster when you teach things in the order a new starter will meet them, though the post calls half the time a target rather than a promise.
- Front loading the whole induction means most of it is taught twice, because people retain what they use in the first week.
- Which drinks matter on a given shift is a data question, answered by the product mix in the point of sale rather than from memory.
- The gap between doing something wrong and finding out is the single biggest determinant of how long onboarding takes.
- A simple record of what each person has been shown stops repeated training and unnoticed gaps, and takes a minute per shift.
The typical induction covers the menu, the till, the recipes, the suppliers, the opening and closing procedures, the licensing rules and the fire exits, across two or three days. The new starter retains a fraction of it, which everyone involved knows and nobody plans around.
The formal part is also short. The Bureau of Labor Statistics describes bartender training as on the job, typically lasting a few weeks under an experienced bartender, so the order in which things are taught in those weeks matters more than the length of the checklist.
The fraction they retain is not random. It is whatever they used in the first week. Everything else has to be taught again later, informally, usually by whoever happens to be next to them.
What should a new bartender learn first?
The single change that compresses onboarding most is teaching things in the order they will be encountered rather than in the order they appear on a checklist.
A new starter on a Thursday evening needs the twelve drinks that make up most of Thursday's orders, the till, and how to ask for help. They do not need the full menu, the stock procedure, or the supplier list, and teaching those on day one guarantees they are taught twice.
Working out which twelve drinks matter is a data question, and most bars answer it from memory. The actual product mix by day is in the point of sale, and it is frequently not what the team assumes.
Teach the balance, not the number
A recipe learned as a quantity is a rule to be followed. A recipe learned as a balance is a skill that transfers.
Someone told a serve is fifty millilitres can make that drink. Someone told it is balanced at fifty, thin at forty and hot at sixty can make that drink, can tell when they have made it wrong, and can reason about a drink they have not been taught. The second takes about a minute longer to teach.
Make the first correction fast
The gap between doing something wrong and finding out is the single biggest determinant of how long onboarding takes. A new starter who drifts for three weeks before anyone mentions it has spent three weeks building a habit that now has to be unbuilt.
This is where the shift level record earns its place. A manager who can see that a particular product drifted on a particular shift can say something concrete two days later. A manager working from a monthly variance figure cannot say anything to a specific person about a specific night, ever.
Give them the reason for the standard
New staff comply with rules they do not understand for exactly as long as they are watched, which is a reasonable response to being handed arbitrary constraints.
The pour standard exists because the drink is correct there and because the margin depends on it. Both halves are true and the first one is the half that survives a busy Saturday. Leading with the margin makes it the owner's rule. Leading with the drink makes it the bartender's standard.
Write down what you taught
Most bars cannot say what a given member of staff has actually been shown, which means training is repeated and gaps persist unnoticed. A simple record of what was covered and when removes both problems, and it takes a minute per shift.
It also means a departure does not reset a new starter's progress, because the next person to work with them can see where they are rather than starting from an impression.
Is half the time realistic for everyone?
Half the time is a target rather than a promise, and it depends heavily on how experienced the person already is. What the sequencing reliably does is reduce the amount of teaching that has to happen twice, which is where most of the waste sits.
CoreTAP supplies the part of this that is a data problem: what actually sells when, and where a specific shift drifted, close enough to the event that somebody can still act on it.
Frequently asked questions
Can I really onboard a bartender in half the time?
The post calls half the time a target rather than a promise, and it depends heavily on how experienced the person already is. What sequencing reliably does is reduce the amount of teaching that has to happen twice, which is where most of the waste sits.
What should a new bartender learn first?
What they will face this week. A new starter on a Thursday evening needs the twelve drinks that make up most of Thursday's orders, the till, and how to ask for help, not the full menu, the stock procedure or the supplier list.
How quickly should I correct a new starter?
As close to the event as you can. A new starter who drifts for three weeks before anyone mentions it has built a habit that now has to be unbuilt, while a shift level record lets a manager say something concrete two days later.
Should I explain the pour standard in terms of margin?
Lead with the drink. The standard exists because the drink is correct there and because the margin depends on it, but leading with the margin makes it the owner's rule, while leading with the drink makes it the bartender's standard.
Why keep a record of what each person was taught?
Most bars cannot say what a given member of staff has been shown, so training is repeated and gaps persist unnoticed. A record removes both problems, and a departure does not reset a new starter's progress.