Each metric answers a different question. Confusing them is how you celebrate the wrong month.
Last updated
Zepth Edge module
Jan 1, 2026
USALI 12th edition became mandatory — and every year-on-year comparison in hospitality breaks this year unless the comparatives are restated
HFTP
Lines moved, schedules changed, new reporting became required. The general lesson outlives the example: whenever a chart of accounts changes, a data-mapping problem arrives disguised as a performance story — and it is very convincing.
3–5 days
the month-end close benchmark. Past a week, fix the process before you argue about the numbers
Practitioner convention
A convention, not a standard. But the logic is hard to escape: commentary that lands on day 12 cannot change the month it describes, and by then it is too late to change the next one either.
4 properties
the minimum competitive set — with at least three unaffiliated, and no single brand above half the supply
STR/CoStar compset guidelines
The general principle is the one worth keeping: a peer group you selected is a peer group you can game. An index of 105 against a compset chosen to produce 105 is a bedtime story with a decimal point.
~$104
GOPPAR in the Middle East — the highest of any region, against roughly $93 in the Americas, $84 in Europe and $52 in Asia-Pacific
HotStats (2024–25 data)
Date-stamped deliberately. Benchmark within your segment and region, never against a global mean — a global average of things that are not alike is a number describing nowhere.
A management-information pack is the monthly performance story an owner actually reads: the standardised operating statement, the KPIs, the benchmark indices, and the variance narrative that explains what happened and why.
Capital gets allocated on the strength of it. Which makes the most dangerous thing in a reporting pack not a wrong number, but a right number answering a question nobody asked.
Every headline metric is blind to something, and the blindness is never randomly distributed — it is almost always blind to cost.
The clearest case is hospitality’s RevPAR, revenue per available room. It ignores every expense and all non-room revenue. So a hotel can grow RevPAR while its actual profit per room falls, and the mechanism is entirely ordinary: growth bought through occupancy drags housekeeping, laundry and a 15–25% distribution commission along with it, while growth bought through rate does not. Same headline, opposite economics. The correctives are TRevPAR, which counts all the revenue, and GOPPAR, which counts what operations actually earn — and GOPPAR is the closest single proxy for what the asset is worth.
The pattern is not hospitality’s. Every industry has a topline metric it celebrates and a contribution metric it should. Revenue that costs more than it earns looks identical, at the top of the page, to revenue that does not. That is the whole problem, and it is why a pack organised around one number is a pack designed to mislead its own readers.
A peer group you chose is a peer group you can game. Benchmark indices — in hospitality, MPI for occupancy share, ARI for pricing power, RGI for overall revenue share — are simply your number divided by the peer group’s, times a hundred. Which means the peer group is doing all the work. The STR discipline is a reasonable floor anywhere: a minimum of four properties, at least three unaffiliated, and no single brand above half the supply. Without something like it, an index of 105 tells you only that somebody chose the comparison well, and the marketing budget it justifies will chase a share gap that does not exist.
The owner and the operator read the same statement and stop at different lines. Both need the operating statement down to gross operating profit. But an owner then cares about what is left after the capital reserve — net operating income, debt cover, asset value — while an operator cares about GOP, its benchmark indices, and whether the incentive-fee hurdle was cleared. One pack must serve both, and USALI’s 12th edition formalises exactly this with a dual summary statement. The failure mode is not disagreement. It is two sets of books, quietly diverging, each honestly maintained.
Close speed is a leading indicator, not an administrative statistic. Three to five working days is the benchmark. The reason is not tidiness: variance commentary that arrives on day 12 cannot change the month it describes, and it is too late to change the following month either — so costs drift for two full cycles before anyone corrects them. If your close takes more than a week, the numbers are not the problem you should be solving.
Data quality is the binding constraint, and it fails in the most expensive possible way. The pack assembles from property, point-of-sale, payroll and back-office systems. When those interfaces are unreconciled, the result is not a gap — a gap would be honest. The result is a FAKE VARIANCE: a number that looks like a performance story, gets explained at length in the meeting, and turns out to have been an interface. Nothing burns the credibility of a reporting pack faster, and once it is burnt every real variance is met with the same shrug.
When the chart of accounts changes, a mapping problem arrives dressed as performance. USALI’s 12th edition became mandatory on 1 January 2026: utilities became an Energy, Water & Waste schedule carrying consumption metrics for ESG reporting; full-time-equivalent reporting by department became required, including outsourced labour; cluster-service costs got formal treatment; and lines moved — waste out of property operations, in-room entertainment into IT. Every year-on-year comparison this year is wrong unless prior-year figures are restated or footnoted. And a variance that is really an account move is the most convincing false signal a pack can produce, because every number in it is correct.
Report the topline alone, and rate-cutting to buy volume looks like success — right up to the point where cost-per-unit and commission load have quietly moved the profit line the other way.
Break the peer group, and marketing spend chases a share gap that was never there. Close slowly, and every decision is made on numbers describing a month you can no longer affect.
And leave the new standard unmapped, and the pack will hand you a set of variances that are not variances at all — with a confident explanation attached to each one, offered in good faith, by people who had no way of knowing.
Bulk ingestion from property, point-of-sale and financial systems, with 50+ hospitality KPIs computed on arrival and USALI-aligned statements produced from them. Occupancy heatmaps, revenue-to-GOP bridges, departmental profitability, and source-market and channel analysis.
Fiscal calendars and currencies are handled at the portfolio level, which is where consolidated reporting usually stops being a system and becomes a spreadsheet with a person attached to it. And every upload carries an AI-generated insight summary — so the pack arrives with its first draft of the story already written.
The pack answers the question that was asked — topline, total revenue capture, and operating profit per unit are reported as the different things they are.
Benchmark indices mean something, because the peer group was constructed to be honest rather than to be flattering.
The close lands inside the window where commentary can still change something.
Interface gaps are caught before the meeting, so a fake variance does not consume the credibility of every real one.
Property, point-of-sale, payroll and financial feeds pulled in together — because a pack assembled by hand is a pack that arrives on day twelve.
Computed on arrival rather than assembled quarterly, with USALI-aligned statements produced from the same data.
Occupancy share, pricing power and revenue share against a compset with enforced integrity rules.
The walk from topline to operating profit, which is where the metric that flatters you gets caught.
By department, source market and channel — so the commission load on volume-led growth is visible rather than inferred.
Across fiscal calendars and currencies, which is precisely where consolidated reporting degrades into a spreadsheet with a person attached.
Daily revenue reconciliation is not an accounting nicety — it is the only thing that makes a three-to-five-day close achievable rather than aspirational.
Three to five working days. If it takes longer, the process is the finding.
Standardised operating statement, the metrics that answer different questions, and benchmark indices against a peer group with integrity.
Rate against volume. Incremental margin against target. Causes, not seasons — and the same decomposition discipline the budget process uses, because it is the same question.
The owner pack, and the portfolio view above it — across fiscal calendars and currencies, which is where consolidation usually breaks.
“Operating profit per room fell 4% on flat topline — driven by energy, water and waste up 18%, and banquet labour.” The pack arrives with its first draft of the story already written, for a human to challenge.
Anomalies across feeds, flagged BEFORE the meeting. An interface gap that gets explained at length as a performance story is the fastest way to destroy a pack’s credibility, and it is entirely preventable.
Properties compared on the same chart of accounts — which, in a year when the standard has just changed, is the difference between a benchmark and an accident.
“Rank properties by incremental margin over the last six months.” Answered from the data, in the form the question was asked.
The engineer’s judgment stays in charge; the AI removes the latency and the blind spots.
The standardised operating statement and its departmental schedules. Topline, total revenue capture and operating profit per available unit — reported as the different questions they are. Benchmark indices against a compset with integrity rules enforced. Incremental margin against budget, with the variance narrative and its causes. The forward look, because a pack that only reports backwards is a history lesson. And the portfolio consolidation across fiscal calendars and currencies.
It became mandatory on 1 January 2026. Utilities became an Energy, Water & Waste schedule carrying consumption metrics for ESG reporting; full-time-equivalent reporting by department became required, including outsourced labour; cluster-service costs received formal treatment; and lines moved — waste out of property operations, in-room entertainment into IT. Restate or footnote your prior-year comparatives, or this year’s year-on-year analysis will confidently report account moves as performance.
Read the full answerAll three, and celebrate the last. RevPAR is the rooms topline and is blind to every cost. TRevPAR captures total revenue. GOPPAR is what operations actually earn per available room, and it is the closest single proxy for what the asset is worth. A hotel can grow RevPAR while GOPPAR falls — that is not an edge case, it is what happens whenever growth is bought through occupancy rather than rate.
Read the full answerYour metric divided by the competitive set’s, times a hundred: MPI for occupancy share, ARI for pricing power, RGI for overall revenue share. Above 100 means you are beating your fair share of the market — but only if the compset is honest. A minimum of four properties, at least three unaffiliated, and no single brand above half the supply. Against a compset chosen to produce a good number, an index of 105 is a bedtime story.
Read the full answerThe standardised summary statement and its departmental schedules, the KPIs and benchmark indices, incremental margin against budget, a variance narrative that names causes rather than seasons, and the forward look — the reforecast. Closed within three to five working days, because a pack that reports only backwards, and late, is a history lesson with a distribution list.
Global all-hotels sits around 38% on 2024–25 data, but the figure moves enormously by segment and region — Middle East GOPPAR leads the world at roughly $104, against about $52 in Asia-Pacific. Benchmark within your segment and your region. A global average of things that are not alike is a number that describes nowhere, and comparing yourself to it tells you nothing you can act on.
Because commentary that arrives on day 12 cannot change the month it describes — and by the time it lands, it is too late to change the next one either. So costs drift for two full cycles before anyone corrects them. Three to five days is achievable with daily reconciliation discipline. Beyond a week, the close is the finding, and no amount of analysis on the far side of it will help.
Related modules
Related answers
Terms defined here
Zepth is the construction project delivery platform — it runs construction, procurement and asset management on one record, and does the work: reading the drawings, reviewing the submittals, matching the invoices and flagging the risks, with a human sign-off on anything consequential.
A short, tailored walkthrough on your real workflow — no generic demo.