Your daily flash report is lying to you

Picture the morning stand-up. The flash report is already in everyone’s inbox, sent overnight by a script or assembled by hand by the night audit. Rooms revenue, occupancy, ADR, food and beverage, labor against cover. The GM reads the top line out loud. Heads nod. The day gets planned off it. Somewhere up the chain a regional or a COO is reading the same numbers across a dozen properties, deciding which buildings to call and which to leave alone.

Now ask the question nobody asks in that room. When was each of those numbers actually true? Not when it was sent. When it was true. The rooms number was pulled at one time. The banquet revenue was pulled at another, before half of last night’s events posted. The labor figure came off a punch export that has not been reconciled against the schedule. The food and beverage line is a snapshot of a POS that was still ringing covers when the report ran. Every number on the page is real. None of them are true as of the same moment, and a few of them are not true at all yet.

That is the quiet problem with the hotel flash report. It is the most-read document in the building and the least examined. We run the whole day off it, and most of the time we never check whether the day it describes is the day that actually happened.

Why the numbers drift

The flash report lies in four predictable ways, and none of them are anyone’s fault.

The first is timing. A flash report is a collage of snapshots taken at different moments. The PMS gets pulled at 2 a.m. so night audit can close. Banquet checks get keyed when the catering office opens, hours later. Labor comes off whenever the punch export runs. The report stitches these together and presents them as one coherent picture of “yesterday,” but yesterday never existed as a single frozen frame. It was assembled from pieces taken at different times, and the seams do not show on the page.

The second is unposted revenue. At a full-service or convention hotel, a meaningful share of last night’s banquet and catering revenue has not posted when the flash runs. The event happened. The money is real. But the check is still sitting in Tripleseat or Delphi waiting to be reconciled and posted, so the flash shows a food and beverage number that is light, sometimes badly light on a heavy banquet night. The GM reads it, sees a soft F&B day, and either worries about the wrong thing or, worse, stops trusting the F&B line entirely and ignores it.

The third is definition drift, and it is the meanest one because it is invisible. The same word means different things in different systems. “Occupancy” might include or exclude house use and comp rooms depending on which export feeds the flash. “Revenue” might be net or gross of the discounts and rebates that post later. Labor “hours” might be scheduled or actual, and the gap between those two is exactly the number a COO is trying to manage. The flash report inherits whatever definition the underlying export happened to use, and that definition was set once, by someone, a while ago, and quietly went stale. This is the same drift that corrupts the handful of PMS fields nobody on the team fully trusts, surfacing one layer up.

The fourth is reconciliation lag. The flash is a pre-accounting document by design. It runs before the controller touches anything. So it carries every keying error, every misposted folio, every comp that should have been a rebate, straight through to the GM’s inbox. Most days the errors are small and wash out. Some days one fat-fingered group master or one missed rebate moves the number enough to change a decision, and nobody finds out until the month closes and the real P&L disagrees with the flash everyone planned around.

Why operators trust it anyway

Knowing all of that, operators still run the building on the flash. We do, and we are right to, and here is why.

The flash report is the only number that arrives in time to matter. The reconciled, audited, true number shows up at month-end, two to three weeks after the day it describes. By then the day is gone. You cannot staff a shift, chase a soft segment, or fix a labor overage on a day that ended eighteen days ago. The flash is fast, and in operations, fast and roughly right beats slow and exactly right almost every time. A COO managing a dozen buildings cannot wait for the accounting close to know which property is sliding. The flash, for all its flaws, is the only thing that tells them today.

So this is not an argument to stop trusting the flash. That would be worse. A GM who ignores the flash and waits for the clean number is a GM who is always managing the past. The argument is narrower and more useful. The flash is going to be the operating number whether it is trustworthy or not, because nothing else is fast enough to compete. So the work is not to replace it. The work is to make the fast number also the honest one.

What a flash report that tells the truth requires

A flash that tells the truth does not need to be slower. It needs three things the typical flash does not have.

It needs one consistent moment. Every number on the page should answer for the same cutoff, and the page should say what that cutoff is, in plain words, at the top. “As of 6 a.m., banquet revenue posted through last night’s events.” When the operator knows the frame, the operator can read the gaps correctly instead of being silently misled by them. An honest flash tells you what it does not yet know. A dishonest one pretends to know everything as of no particular time.

It needs reconciled definitions, written down and held still. Occupancy means one thing. Revenue means one thing. Labor hours means one thing, and the same thing tomorrow, and the same thing across every property the COO compares. This is not a technical problem. It is a governance problem, the discipline of writing the definition once, in language an operator can read, and refusing to let the export quietly change it. When two buildings define occupancy differently, the COO is not comparing two buildings. They are comparing two dictionaries and calling it a portfolio.

It needs its sources to actually connect. The reason the flash is a collage of mismatched snapshots is that the systems underneath it never join. The PMS, the sales-and-catering system, the POS, and the labor platform each hold a true piece, and the flash is whatever a script or a night auditor could staple together before the deadline. The truthful flash is the one built on a connected source where the banquet check, the room revenue, and the labor punch live in the same place, reconciled against each other, before the report runs. That is the same root cause underneath nearly every reporting failure we see, the one where the building runs on five systems that disagree and calls the disagreement a report.

This is the operating cadence part of the Operator Intelligence model. One connected source of truth so the numbers come from the same place. An intelligence layer that reads the way an operator already thinks, so the flash answers the questions a GM actually asks. And a cadence that makes a trustworthy flash the daily habit, not a clean number that arrives too late to use. The flash report is where that cadence lives or dies, because it is the one document the whole building already reads every single day. Fix the flash and you have not fixed a report. You have fixed the number the building runs on.

You can do this yourself. Pin the cutoff, write the definitions down, hold them still, and push to post banquet revenue before the flash runs instead of after. Those moves cost no software and they make tomorrow’s flash more honest than today’s. If you would rather the honest flash just arrived every morning, reconciled and consistent, without a night auditor heroically stapling exports together at 2 a.m., that is the kind of thing we build. Either way, stop reading the flash as gospel and start reading it for what it is. A fast, useful, slightly dishonest first draft of the day, that is worth making true.