Loading
Loading
An independent Houston bar with a loyal room and a flat trend. No obvious problem, which is the hardest kind to fix, because there is nothing to point at.
This venue was not in trouble. It had regulars, a decent reputation, and an owner who had been behind the bar himself for years and knew the business. Revenue was stable. Costs were roughly where he expected. Nothing was on fire.
That is a harder starting point than a crisis. When something is visibly broken you at least know where to aim. When the business is merely fine, every decision is made on instinct, and instinct in hospitality is a mix of genuine expertise and habits that stopped being true two years ago. Nobody can tell which is which from the inside.
The question he actually wanted answered was simple and unanswerable with what he had: of the things we do, which ones are working?
It is worth saying this part first, because it is the part most case studies skip.
He believed his Thursday crowd was his best crowd. It was. He believed a specific member of staff was carrying the busiest shifts. She was, by a clear margin, and the data made a raise easy to justify. He believed his cocktail list outperformed his beer program on margin. Correct.
An operator with real floor experience is right about most things. If the data had contradicted him constantly, the sensible response would have been to distrust the data. What made the exceptions credible was that the rest of it confirmed what he already knew.
A long-running weeknight promotion was widely understood to bring people in. It did bring people in. Broken out properly, those people arrived, took the discounted item, and left, with an average check well below a normal night on the same weekday. The night looked busy. It was busy. It was also, on contribution, one of the weakest nights of the week.
Busy and profitable had been treated as the same signal for years, because from behind the bar they feel identical.
Comps were running higher than anyone had estimated, and they were concentrated. Not theft. A friendly bar where buying a round for a regular had become the default way to handle almost anything, and where nobody had ever seen the total.
The fix was not a crackdown. It was showing the team what it added up to over a month. Most of it stopped on its own once it was visible, which is generally how this goes when the cause is habit rather than intent.
Both were popular. Both had been priced when the ingredient costs were different. Neither had been revisited, because there is no natural moment at which anyone recalculates the margin on a drink that sells well. Popularity had been standing in for profitability.
The honest framing here is not a headline multiple. It is that the comp leakage alone, once it was visible, was worth more per month than the platform cost. Everything after that was upside.
That is the usual shape of the return in a venue this size, and it is worth stating plainly because software gets sold on transformation and mostly delivers on attention. Nothing here required a new concept or a renovation. It required knowing which of four plausible explanations was the true one, which is exactly the thing no amount of experience behind the bar can supply on its own.
The bar still has a slow stretch in the middle of the week that no amount of scheduling solves, because the neighborhood is quiet then. Analytics does not create demand that is not there. It stops you spending money pretending otherwise, and it tells you which of your existing nights has room left in it.
Most of them, probably. The value is in finding the two that are not. Bring two weeks of data.