Did your happy hour actually work?
"It's so difficult for me to buy into any of that because you will never ever correlate that to Sales." That's an operator on marketing spend, and he's right about attribution — nobody asks every guest why they came. But you don't need attribution to answer the smaller, more useful question: did the window with the promo in it ring more money than the window without?
That's two numbers from your POS for each of two weeks, and subtraction. Five minutes. Here's the sheet.
What to pull first
- Pick the same window on both sides. Same weekday, same hours, same length. Comparing a promo Friday to an ordinary Tuesday measures the day of the week, not the promo.
- Covers, not checks. Bodies through the door in that window. If your POS counts guests, use that; if it counts tickets, use tickets on both sides and say so to yourself.
- Average check as rung. What guests actually paid, discount already applied — not menu price.
- Use several weeks if you have them. One week against one week is a coin flip with weather in it. Four against four is an answer.
The numbers
Leave it blank if you'd rather not guess. Blank means unmeasured, not zero.
Nothing you type here leaves your browser. The math runs on this page, we don't send it anywhere and we don't store it.
How to read the answer
- A fuller room is not a better week. The most common way a happy hour "works" and loses money is more covers at a smaller check. That's why the revenue line sits under the covers line here, not beside it.
- Break-even is the honest bar. If the promo window needs forty more covers to ring what the ordinary window rang, forty is the number the discount has to buy before anything good starts happening.
- Watch who came back. One owner ran 50% off and reported: "alot of people came but not the customers that are supposed to be our regulars." Run the same check on the four weeks after the promo ended. If the baseline went back exactly where it started, you rented traffic rather than building it.
- Don't grade it on likes. The trade-press warning is that "people tend to follow vanity metrics like likes and views". Covers and check are the two numbers that show up in the bank.
What this can't tell you
This is revenue, not profit. It has no food cost, no pour cost, no labour and no product mix, so it cannot tell you whether the extra covers were profitable — only whether they were bigger. Your POS knows the half we don't: which items moved, what they cost you, whether the promo cannibalised your full-price dinner or filled a genuinely empty room. It also can't establish cause. A promo week that rang better may have had better weather, a game on, or a neighbour closed for repairs; the calculator measures the difference and stays quiet about why. Treat a single week as a hint and four weeks as evidence.
Your block, not the city.
This page gives you one median per neighborhood. Your report gives you the venues within walking distance of your door — what each one charges, item by item, and where your menu sits against them.