Loading
Loading
Operators expect to find one big leak. There is almost never one big leak. There are forty small ones, none of which would survive being looked at directly, all of which have survived for years because nobody could look at them at all.
Every individual instance of loss in a bar has a reasonable explanation. That is the whole problem.
The pour was heavy because it was a regular. The void was rung because the guest changed their mind. The comp was because the wait was long. The discount was applied because the manager said it was fine. Each of those, examined alone, is defensible. Several are genuinely correct.
Because every instance is explainable, nobody escalates any instance. And because the total is never assembled anywhere, the sum of forty defensible decisions never gets examined at all. It shows up as a cost percentage that is slightly worse than it should be, which gets attributed to supplier pricing.
The single largest and most consistently underestimated line. A quarter ounce of overpour is imperceptible in a glass and it is not imperceptible across a keg. It also does not feel like loss to the person doing it. It feels like hospitality, and to be fair, it partly is.
What makes it tractable is measurement rather than instruction. Telling a team to pour accurately achieves very little. Showing a bartender that their pour is running consistently heavier than the rest of the bar, with a number attached, tends to correct it within a week without anyone needing to be disciplined.
Comping is a legitimate service tool that drifts into being the default answer to every situation, because it is the fastest way to end an awkward moment and it costs the person doing it nothing.
The correction is almost never enforcement. It is visibility. Most teams shown their monthly comp total reduce it themselves, because nobody intended the number to be that size. They were each making one small decision at a time.
Individually explainable, in aggregate a pattern, and one of the few signals here that is worth treating as potentially deliberate. Late voids on already-settled checks concentrated on one operator is not a training issue.
The reason this is worth naming carefully is that the pattern, not the incident, is the evidence. Confronting somebody about a single void gets an explanation you cannot disprove. Presenting eleven weeks of a rate four times the bar average is a different conversation and a much shorter one.
A short delivery accepted at the door is unrecoverable by the time the count happens. This one is pure process, it is boring, and it is worth real money in venues that receive frequently.
Not loss exactly, but it behaves identically on the P&L. A drink priced when the ingredients cost less and never revisited, selling well, quietly negative. Popularity is regularly mistaken for profitability, and the better it sells the less anyone questions it.
This is the part operators find hardest to believe and it is the least impressive fact about it. Nothing new is being collected. Every void, comp, discount and pour was already in the POS. The information was sitting there the entire time in a form nobody could interrogate.
That is why the first read is usually same day rather than the end of a baseline period. There is no learning window required to tell you what already happened last month.
The clean way to state the return is this: in most venues, the recoverable leakage identified in the first reporting cycle exceeds the cost of the platform. Not over a year. In the first look.
That claim sounds aggressive and is actually a fairly low bar, because the comparison is against a subscription and the leakage in a venue that has never measured pour variance is typically substantial. It says more about how invisible the losses were than about the software.
What we will not claim is a specific multiple. Any vendor quoting you a precise ROI figure before seeing your data is quoting you their average customer, and the range across venues is enormous. A tightly run bar that already measures variance may find very little. That is a good outcome and it is also a real possibility, and you should be told so before you buy rather than after.
Finding leakage once is a one-off saving. The durable value is that the number now has an owner and a weekly rhythm.
Cost control decays. A pour standard drifts as staff turn over. Comp rates creep back up after a busy quarter. A supplier price rises and the menu does not move. None of that is negligence; it is entropy, and it is why the venues that hold their gains are the ones where somebody looks at the same four numbers every week, not the ones that ran a project.
This is also where turnover and cost control turn out to be the same problem. A venue that replaces its bar team twice a year is re-teaching the pour standard twice a year and never reaching the point where it holds on its own. That case study is here.
Bring two weeks of data. If there is nothing meaningful to find, we will tell you that.