Loading
Loading
A hotel bar in Houston was running between $45,000 and $50,000 every two weeks. It now runs between $80,000 and $90,000. The room did not change. What changed was how much of it anyone could see.
A hotel bar is a strange business. It has guaranteed footfall and almost no control over it. Guests arrive because of the hotel, not because of the bar, which sounds like an advantage and mostly functions as an excuse. When revenue is flat, it is very easy to conclude that occupancy was flat and there is nothing to be done.
That was roughly the position here. Two week revenue sat in a band between $45,000 and $50,000 and had for a long time. Nobody thought the bar was broken. There was no crisis. It was simply assumed that the number was the number, because the number was a function of the hotel.
The reason that assumption survived is that nothing available to the operator could test it. The property reported revenue. It did not report what happened inside the revenue.
The first useful thing was not a single dramatic discovery. It was that the bar had been treating every hour as broadly similar, when the transaction data described a room with a very sharp shape.
Check-in traffic, the pre-dinner window, and the late return from downtown produced three distinct peaks. Staffing had been built around a flat evening shift, so two of those peaks were served by a bar that was short, and the trough in between was served by a bar that was overstaffed.
Short-staffed peaks do not usually show up as lost revenue in any report. They show up as guests who looked at the queue, decided against a second drink, and went upstairs. That is invisible on a P&L and enormous in aggregate.
The menu had been built to look complete rather than to match who was actually standing there. A business traveler ordering alone before dinner is a different customer from a group returning at eleven, and the bar was offering both the same list. Once the mix was broken out by daypart, several items turned out to exist almost entirely on the menu rather than in the sales data.
Average check varied more between individual bartenders than it did between weekdays. That is a strong signal, because the room is broadly the same on a Tuesday regardless of who is behind the bar. When the person changes the result more than the day does, the difference is not demand.
None of the changes were dramatic, and that is worth being honest about. There was no renovation and no rebrand.
Revenue moved from the $45,000 to $50,000 band up to $80,000 to $90,000 per two weeks. It is worth being precise about what that is and is not.
It is not a new market. The hotel did not get busier. It is the same building, the same guests, and broadly the same hours.
What it represents is the distance between what the room was already capable of and what it was actually capturing. The demand had been walking through the lobby the entire time. The bar simply was not organized to meet it, and no report available to the operator was capable of pointing that out.
The increase held because none of the changes depended on sustained heroics. A schedule that matches demand does not require anyone to work harder. A shorter menu is easier to execute, not harder. Consistency between bartenders reduces effort once it is established. If the gains had rested on somebody caring more than usual, they would have decayed the moment that person had a bad month.
The specifics here belong to this venue. The pattern does not.
Most rooms are under-capturing rather than under-visited, and the reason it persists is that flat revenue is one of the easiest things in hospitality to explain away. Occupancy, weather, the economy, the season. Every one of those is sometimes true, which is exactly what makes the explanation so durable and so expensive.
The only way to know is to look inside the revenue rather than at it.
Bring two weeks of data. We will show you the shape of your demand curve against the shape of your schedule.