On this page
Key takeaways
- Running one standard across several bars means writing it down, because at one site the standard was usually a person, not a document.
- The durable failure is not a worse second site but two acceptable sites that have quietly become different businesses.
- Define the worst acceptable version of a shift first: sites can differ above that floor but not below it.
- With two sites, the difference between them is the most useful information, since each location is a control for the other.
- Route each finding to a named role at the site, not the group office, and check the following week whether it moved.
A single good bar is usually run by proximity. The owner or the general manager is in the room often enough to see what is happening and correct it in the moment. That works, and it works well enough that the standard never has to be made explicit. It lives in one person's judgement and gets applied continuously.
Open a second site and that mechanism stops working, without any decision being made to stop using it. The owner is now in each room half as often. The standard has not changed, but the thing that was enforcing it has been halved.
The failure mode is not the one people expect
Groups usually brace for the second site being worse than the first. Often it is, at the start, and that is treated as a training problem that will resolve.
The more durable failure is different. Both sites end up acceptable, and different. Each develops its own version of the pour, the handover, the response to a complaint, the way a Friday is staffed. Neither is bad. They are simply not the same business, and a guest who uses both notices immediately even if they could not articulate what changed.
At that point the brand is no longer a promise about what happens when you walk in. It is a logo shared by two operations, and the value of the second site to the first is close to zero.
What has to become explicit?
The uncomfortable part of multi site operation is that everything previously held in one person's head has to be written down, and writing it down reveals how much of it was never really agreed.
Three things matter most, in this order.
The floor, not the ceiling
Do not start by standardising the good nights. Start by defining the worst acceptable version of a shift: the pour, the wait, the handover, the response when something goes wrong. Sites can differ above that line. They cannot differ below it. This is a much easier thing to get agreement on than a definition of excellence, and it is the part guests actually notice.
The comparison
The moment there are two sites, the most useful piece of information in the business is the difference between them. Same product, same night of the week, two different variance figures. That comparison is where a group has an advantage a single site does not: each location is a control for the other.
Most groups cannot make that comparison, because each site's numbers arrive in a different shape at a different time, and reconciling them by hand is a job nobody has. That is a tooling problem rather than a management problem, and it is worth naming as one.
The routing
A finding that reaches the group office and stops there has not changed anything. The question that matters is who at the site is going to do something about it, and by when. Multi site operations fail at this point more often than at any other, because distance makes it easy for everyone to assume someone else has it.
What does not work
Two approaches are common and both tend to disappoint.
Flying the good manager between sites. This works while it is happening and stops the moment it stops. It also burns the person doing it, who is usually the person the business can least afford to lose.
A weekly group report. Reports address nobody. A pack that lands on Monday summarising last week gets read, agreed with, and does not produce an action, because no line in it is any specific person's job.
The version that does work
Define the floor. Measure at the level of the shift so sites are comparable. Route the finding to a named role at the site, not to the group. Check the following week whether it moved.
That is unglamorous, and it is roughly what CoreTAP automates: reading each site's point of sale, working out what the numbers mean while there is still time to act, and handing the next move to the role that owns it. If you would rather compare the tools in the category first, the comparisons are the honest version of that.
Frequently asked questions
Why did my standards slip when we opened a second bar?
A single bar is usually run by proximity: the owner or general manager is in the room often enough to correct things in the moment. Open a second site and that person is in each room half as often, so the thing enforcing the standard has been halved.
Where should we start standardising across sites?
Start with the floor, not the ceiling. Define the worst acceptable version of the pour, the wait, the handover and the response when something goes wrong, which is easier to agree than a definition of excellence and is what guests notice.
Can't I just move my best manager between locations?
This works while it is happening and stops the moment it stops. It also burns the person doing it, who is usually the person the business can least afford to lose.
Is a weekly group report enough to keep sites aligned?
Usually not. A pack that lands on Monday summarising last week gets read and agreed with but does not produce an action, because no line in it is any specific person's job.
Why is it so hard to compare our sites?
Each site's numbers often arrive in a different shape at a different time, and reconciling them by hand is a job nobody has. That is a tooling problem rather than a management problem.