Most Hotel Budgets Are Agreed. Few Are Owned.

The growth line is usually the least owned line in the document.

Somewhere, a director of sales is sitting across from a general manager, and the number on the screen is last year plus eleven per cent. Everyone in the room knows how it was built. Everyone will sign it. And by June, when it is running four points behind, the conversation will be about effort, because effort is the only thing left to argue about when nobody can say where the number came from.

I have been in that room from both sides of the table, and in a few places in between. The pattern barely changes. The budget is anchored to last year's actual, adjusted for whatever the market is expected to do, pushed up by the management's expectations and then reverse-engineered into segment targets so that the columns add up. It is a legitimate way to produce a spreadsheet. It is not a way to produce a plan.

The question nobody asks in September

McKinsey has been writing about this for over a decade. Lovallo and Sibony showed how budgets across industries are anchored to the prior year and adjusted only at the margins, so that the final number looks remarkably like the one managers proposed at the start of the process. That is a fact about human beings, not hotels. But hotels have a particular version of it, because our budgets are approved at segment level, and a segment is exactly the right size for speculation to hide in.

You can add 8% to corporate without naming a single company. You can add fifteen per cent to groups without a lead in the system. The number survives the meeting because it is plausible and because the alternative, a lower number, is a harder conversation with the owner than a miss will be next July.

The question that breaks this is simple and almost never asked: of next year's revenue budget, what percentage can we name today, by account?

What I found when I asked it

A few years ago I ran that exercise on a city hotel whose budget had just been approved with a healthy double-digit uplift. The composite numbers here are illustrative of the pattern rather than any single property's figures, but the shape is one I have seen repeatedly.

Signed contracts with committed volume and dates, the crew contracts and long-stay agreements, made up roughly a fifth of the number. Accounts with two or more years of consistent production and a live relationship, but no commitment, made up another forty per cent. Named opportunities actually in the sales cycle, RFPs in play and groups with dates, added perhaps fifteen. The remaining quarter of the budget could not be attached to anything. It was market growth, a new source market somebody had mentioned in a meeting, and a rounding-up.

That was uncomfortable enough. The genuinely alarming part was what happened when I looked only at the growth increment. The uplift the owner had approved, the eleven per cent everyone had congratulated themselves on, sat almost entirely in that unattached quarter. The base was owned. The growth was a wish. And the document treated both with the same font.

The reframe: certainty, not accuracy

The industry's reflex when budgets miss is to buy better forecasting. Pace tools, BI dashboards, a revenue system with more sophisticated demand modelling. Those tools are worth having, and I would not run a hotel without them. But they answer a different question. They forecast demand. They do not tell you who in the building owns the number, or how much of it exists outside anyone's pipeline.

Budget accuracy is a forecasting problem. Budget ownership is a provenance problem. Most hotels have the first one under reasonable control and have never measured the second.

The Revenue Provenance Ladder

So I started tiering budgets by certainty, in four rungs.

Contracted. Signed commitments with volume and dates. Crew contracts, long-stay agreements, confirmed group blocks, government framework agreements with committed nights.

Repeatable. Accounts that have produced consistently for two or more years and where the relationship is alive today, but where nothing is committed. Negotiated corporate rates, consortia, the travel management accounts that always come back.

Pipelined. Named opportunities at a stage of the sales cycle. An RFP in play, a tender pending, a group lead with dates and a decision-maker attached.

Speculative. Everything else. Market growth, new segments, source markets not yet worked, and any part of the number that cannot be attached to an account.

Two ratios fall out of this. The Provenance Ratio is contracted plus repeatable, divided by the total. It is the share of the year that is real on the day the budget is signed. The Generation Load is the speculative share, and it is the honest measure of how much of the number the team has to create rather than collect.

The decision rule I use is that once the Generation Load passes thirty per cent, the document in front of the owner is no longer a budget. It is a demand-generation mandate wearing a budget's clothes, and it should be approved as one: with the headcount, the source-market travel and the marketing spend that generating that demand will cost, or with a smaller number.

How to use it this month

The audit takes an afternoon, not a consultancy. Take the rooms revenue budget by segment and force it down to account level. Tag every account or block into one of the four rungs. Anything that cannot be tagged is speculative by definition; do not let the team argue it into Repeatable because the market "feels strong". Sum the rungs, calculate the two ratios, and then run the same exercise on the growth increment alone. That second calculation is the one that changes the meeting.

Then take the ladder into the room with the owner. Not as an excuse for a lower number but as the composition of whichever number is approved. In my experience owners rarely reject a number. They reject a number that cannot explain itself, and they remember the ones that did.

Finally, keep tracking it. Certainty decays. A contract lapses and slides to Repeatable. A Repeatable account changes procurement lead and slides to Pipelined. The month-by-month movement of the Provenance Ratio, what I think of as provenance drift, is the early-warning signal most hotels currently receive only when the month closes.

What an unowned budget costs

Three things, and none of them show up as a line item.

The first is unfunded generation. When a quarter of the number is speculative and nobody labelled it, the team is being asked to generate demand with a harvesting team's resources. Most cannot, and the shortfall is then blamed on the people rather than the design.

The second is rate. An unowned number gets chased in the third quarter with the only lever that moves quickly, which is discount. Every hotel that has watched its ADR erode in September while occupancy holds has seen a provenance problem arrive disguised as a pricing decision.

The third is trust. An owner who approved eleven per cent and received seven does not hear "the market softened". They hear that the commercial team did not know its own number. That is slower to rebuild than any rate, and it follows a director of sales to the next property.

The principle

Agreeing a number is not owning it. A budget is only as strong as the share of it you can name, and the growth line is the part most worth naming, because it is the part everyone will remember.

Run the ladder on your budget before it is signed. If you cannot, run it the week after, because the owner will eventually run a version of it for you.

Revenue follows clarity.

Sources for the McKinsey references

• Dan Lovallo and Olivier Sibony, "Is your budget process stuck on last year's numbers?", McKinsey Quarterly, 1 March 2014: https://www.mckinsey.com/capabilities/strategy-and-corporate-finance/our-insights/is-your-budget-process-stuck-on-last-years-numbers

• McKinsey, "How to be objective about budgets", Inside the Strategy Room podcast with Tim Koller and Dan Lovallo, 7 March 2019: https://www.mckinsey.com/capabilities/strategy-and-corporate-finance/our-insights/how-to-be-objective-about-budgets