Hotel Procurement Is Broken. Here's What Rebuilding It Looks Like.

Hotel procurement is the last part of our industry nobody wants to talk about.

• Three quotes by email.
• A spreadsheet someone built in 2019.
• A vendor your predecessor liked.
• A PO that gets approved because it is Tuesday.

We have all done it. We are still doing it.

Why it stayed broken.

It is not because hoteliers are careless. It is because procurement sits between two departments that each consider it somebody else’s job. Operations orders. Accounting pays. The gap between them is where the process was supposed to live, and mostly there is nobody standing in it.

It also never got the software. Every other function in a hotel has a system built for hospitality — the PMS, the revenue system, the POS, the maintenance system. Procurement got a shared drive and an email folder. The tools that do exist were built for manufacturers and distributors, and they assume a catalog, a warehouse, and a purchasing department of six.

A select-service property has none of those. It has a GM, a controller who also covers two other hotels, and an executive housekeeper ordering from a rep she has known for nine years.

The three-quotes ritual.

The three-quote rule is the clearest example of the problem. It exists to prove diligence, and in practice it mostly proves email. Two of the quotes come from suppliers nobody intends to use, gathered so that the third one can be approved.

The rule is not wrong. It is just the wrong unit of work. Diligence at the moment of purchase does nothing about the eleven months afterward, when the agreed price drifts and nobody re-quotes anything.

The quotes rarely survive contact with the invoice either. A price agreed in March shows up differently in September, and nothing in the three-quote ritual was ever designed to notice that.

The 2019 spreadsheet is not the villain.

I want to defend the spreadsheet for a moment, because it usually takes the blame. Somebody built it to solve a real problem, and for a while it genuinely worked. What fails is not the file. What fails is that it only knows what somebody typed into it, and the typing stopped when that person moved to another property.

Every manual procurement tool has the same failure mode. It captures a decision at one moment and then drifts away from reality, while the invoices — which never stop arriving — keep telling the truth to nobody.

Where the record already is.

Here is the thing I did not expect when we started building. The hotel already has the data. It is in the invoices. Every price you agreed to, every quantity you received, every supplier you use, every increase you absorbed — all of it has already been printed, delivered, and filed in an inbox.

Nobody needs to build a procurement history from scratch. Every hotel already has one. It is just stored as paper.

That reframed the whole problem for us. Rebuilding hotel procurement does not have to start by convincing a property to adopt a new ordering process. It can start by reading what the property is already receiving.

What rebuilding it looks like.

So the first thing Atrium does is the least ambitious thing possible. Invoices get forwarded to an email address. Nothing is installed, nothing is integrated, nothing is migrated. Forward an email — that is the whole setup. From there, the process gets its missing pieces back.

• What was ordered, what arrived, and what was invoiced, compared before payment rather than after.
• Every recurring item carrying its own price history, so drift shows up on the line instead of in somebody’s memory.
• Supplier pricing agreements checked against the invoices that are supposed to follow them.
• Comparable suppliers priced side by side, with pack sizes normalized so the comparison is honest.
• Department budgets and spend trends per property, plus a rollup across every property in a group.

It is not a marketplace and it is not an accounting system. It is the review that should have been happening between the order and the payment, and mostly was not.

What it turned up at one property.

At the Niagara Falls hotel we are piloting with, eight months of ongoing use across 137 invoices and 1,214 line items has surfaced 24 price increases averaging 12.3%, 24 comparable-supplier opportunities worth about $796 a month, and 10 spend anomalies. None of it was hidden. All of it was on paper somebody had already approved.

I do not think that property is unusual. I think it is the first one where anyone bothered to line the invoices up.

The decisions stay yours.

None of this decides anything. Atrium does not move money, does not pay, and does not dispute. It flags what is worth reviewing and shows the evidence beside it. Whether a 12% increase is worth a fight, whether a rep gets one more chance, whether the cheaper supplier can actually deliver on a holiday weekend — those are judgment calls, and they belong to the person who knows the property.

That is the part of hotel procurement that was never broken. It was just buried under the part that was.

If you run a hotel, I would genuinely like to hear it: what is the worst procurement story you have got? And if you would rather see it than tell it, hotel procurement software with a pre-payment exception review built in starts with about ten invoices and costs nothing to try.