Use case: measuring the real impact of a promotion

A hotel runs a promotion, two nights for the price of one and a half, a weekend with breakfast included, a flash rate to fill a dead Tuesday, and a few weeks later someone in the meeting says the magic words: “that promo worked, let’s do it again.” Nobody argues. It gets repeated. And right there, in that comfortable nod, lives one of the most expensive and most invisible mistakes in hotel operations: confusing movement with results. The room got booked, yes. But did it leave money behind, or just activity? The thesis of this case is uncomfortable and simple: a promotion without measurement isn’t a strategy, it’s an expensive hunch.

The symptom: the promo repeated “because people liked it”

The symptom is easy to spot because we have all lived it. A promotion launches, occupancy on those days rises, the team celebrates, and the decision to repeat it gets made on an emotional data point, “we filled up,” “guests asked for it,” “it felt busy”, instead of a financial one. The feeling of success is real: seeing the hotel full produces relief and pride. The problem is that “full” and “profitable” are not the same word, and sometimes they are not even relatives.

What happens is that the promotion gets judged by its most visible effect (how many bookings came in) and never by its real effect (how much margin was left after discounting, after paying the channel commission, and after subtracting the sales that would have happened anyway). The visible effect is loud and looks pretty. The real effect is quiet, and it only shows up if someone goes looking for it in the numbers.

The problem: what the hunch never sees

When a promotion is judged “by ear,” three things are systematically left out of the analysis. Not because the team is careless, but because they are things you cannot see at a glance: you have to cross tables to find them.

What sold was never crossed with its cost

A booking under a promotion was, by definition, sold cheaper. If nobody compares the promotional rate against the cost of serving that room, cleaning, amenities, the breakfast you gave away, energy, the commission of the channel it came through, then the “success” is a top-line number (gross revenue) hiding a bottom-line number (net margin). Selling a lot at a negative margin isn’t good business; it’s funding someone else’s vacation.

Cannibalization went unseen

This is the subtlest trap. Some of the bookings that came in “thanks to” the promotion would have come in anyway, at full rate, with no discount at all. The promotion didn’t attract those guests: it handed them money they were already going to spend. That’s called cannibalization, and it’s the difference between winning new customers and simply making the stay cheaper for the regulars. A promo can look spectacular on occupancy and, at the same time, be destroying revenue you already had locked in.

It was never compared against a baseline

Without a baseline period, the same weekend last year, the two prior weeks with no promo, a set of comparable dates, there is no way to know whether the bump came from the promotion or from the season, from an event in town, from the weather, or from pure chance. “It went up” means nothing if you do not know what it went up against.

Selling a lot isn’t the same as selling well. Occupancy fills the hotel; margin fills the till. And only one of the two makes payroll.Hotel operations principle

The method: how you actually measure a promotion

Measuring a promotion’s impact does not require a PhD in statistics. It requires crossing the sources you already have, bookings, rates, channels, payments, guests, into a single structure and asking them four honest questions. Spider Data exists precisely for this: it brings together the eight sources of your operation (bookings, cash, channels, payments, guests, orders, shifts, cash movements) and lets you build the cross without writing code, dragging fields in plain language. It won’t tell you what price to charge, that’s not its job; it tells you what happened and why.

Step 1: isolate what sold under the promo

First you need to be able to separate, out of all bookings, exactly which ones came in under the promotional rate. This is a cross between the bookings table and the rate or promotion applied. Without this separation, everything else is impossible: you would be measuring a promo blended with normal business, like weighing a suitcase without taking it off your shoulder.

Step 2: compare against a baseline period

Take an equivalent period without the promotion, ideally comparable in seasonality, and set it side by side. Here is where calculated fields come in that you don’t have to program by hand: ADR (the average rate per night, which in plain terms is simply how much you charged on average for each room-night), the number of nights sold, and the lead time the booking came in with (how many days before arrival the guest booked). If the promotion worked, it should show up in more than volume: in sustained ADR, in earlier bookings, in longer stays.

Step 3: look at margin, not gross revenue

This is where most analyses stop too soon. Adding up what came in is not enough. You have to subtract the cost of serving those rooms and the channel commission. What remains is the net margin, and that is the only number that truly matters. A promotion that lifts gross revenue 30% and net margin 2% is not a win: it’s nearly-free labor.

Step 4: ask whether it brought new people

Cross the promotion’s bookings with the guests table. How many of those names are customers who had never come before, and how many are repeat guests who already love you and to whom you just handed a discount they didn’t need in order to return? A promotion that draws new faces has a value that goes beyond those nights, it can turn into guests who come back at full rate. A promotion that only discounts the regulars is a slow leak.

The result: money, or just movement

Let’s walk through a purely illustrative example, so the mechanics are clear (the numbers are made up; suppose a hypothetical scenario). Imagine a weekend promotion generated 40 bookings, against 25 in a comparable baseline weekend. At first glance: great success, 15 extra bookings. But once you cross it with cost and the baseline period, another story appears: suppose that of those 40, around 22 would have come in anyway at full rate (cannibalization), the average discount was aggressive, and the channel commission on the new arrivals ate a good chunk of the margin. The illustrative result: the promo moved the hotel, yes, but the weekend’s net margin came in lower than the baseline period without a promotion. Movement, yes; money, no.

Or the opposite can happen, and that is exactly the valuable part: maybe the promotion brought 12 genuinely new guests, on dates that were historically empty, at a margin that, even with the discount, stayed positive, and three of those names have since returned at full rate. In that case the promo didn’t just leave movement: it planted customers. The difference between the two scenarios cannot be seen “by ear.” It only appears when you cross what sold with its cost, against a baseline, looking at margin and at guests.

And because the data in Spider Data is live, not last night’s close, you don’t have to wait until the end of the quarter to know whether a promotion is leaving money or just noise. You see it while it runs, with dashboards that filter each other, and you can schedule the report to arrive on its own every Monday, or set an alert to warn you if the promo’s margin drops below a threshold. If your analyst prefers Power BI, Tableau, or Looker, the numbers come out through an open connection with a token, it’s not a cage; and with R2-Index you can contrast your performance against a reference index, to know whether your “good weekend” was good only for you or genuinely good.

What to measure about a promotion

If you are going to run a promotion and you want to truly know whether it worked, these are the questions your data should be able to answer before you repeat it:

• Bookings under the promotion vs. a comparable baseline period (not against zero).
• Effective ADR of the promo vs. normal ADR: how much you actually dropped the price.
• Net margin per booking: revenue minus cost to serve minus channel commission.
• Estimated cannibalization: how many of those bookings would have come in at full rate anyway.
• New guests vs. repeat guests: did you draw new faces or discount the regulars?
• Lead time: did the promo pull bookings forward, or just shift the date of ones already coming?
• Channels it came through: did it arrive via your own site (no commission) or a channel that took the margin?
• Tail effect: did any of those new guests return later at full rate?

Deciding better, not deciding blind

None of these questions takes the instinct out of operations. The hotelier’s nose, reading a season, sensing which date needs a push, is irreplaceable, and this case does not aim to replace it. What it proposes is far more modest and far more powerful: putting numbers to the instinct before repeating a decision that costs money. The difference between a good hotelier and an exceptional one is rarely intuition; it’s the discipline of checking it.

A promotion you measure teaches you something even when it goes wrong: it tells you which discount was too deep, which channel ate the margin, which dates respond and which don’t. A promotion you don’t measure leaves you only a feeling, and feelings cannot be improved next year. Repeating it “because people liked it” is repeating blind. And when what’s at stake is the margin of your business, there’s a line worth nailing to the office wall: a promotion you don’t measure isn’t a strategy, it’s a bet.