Hotel Flow Through: Why More Resort Revenue Is Not More Profit

Hotel flow through is the share of each new revenue dollar still there as profit once you have paid to earn it. In 2024 it turned negative at US hotels: every extra $1 of revenue came with $1.04 of extra cost, $0.75 of it labor (HotStats, a benchmark owned by Duetto, 2025) [1]^[2]. That is how a resort has its best revenue year and keeps less. The fix is revenue that costs little to deliver, judged by what each department keeps.

Departure day with young kids has its own arithmetic. The flight is at 6pm, everyone needs a nap, and checkout is 11am. I would pay for a 4pm checkout without blinking. I am a guest, not an accountant, but the numbers taught me something: what I most want to buy costs the resort very little to sell, if someone has planned housekeeping around it.

The In-Stay Revenue Report from ResortConcierge AI Research pulls together the published evidence on where resort money comes from and how much reaches profit. This post puts its profit chapter in plain English.

If software is part of your plan, the guide on how to evaluate a hotel AI concierge covers what to check. The arithmetic comes first.

What is hotel flow through, in plain English?

CoStar's definition: "Flow Through measures how effectively incremental revenue is retained as profit after operating costs" (2026) ^[3]. The formula is the change in gross operating profit (GOP) divided by the change in total revenue, times 100 ^[3].

GOP is total operating revenue minus total operating expenses (Mews, a vendor, 2026) ^[4]. In owner words, it is what is left before property tax, insurance, management fees and debt ^[4], or in EHL Insights' phrase, "operating efficiency before considering fixed charges" (2025) ^[5]. Flex is the mirror image: how well a hotel protects its margins "when revenue decreases" (CoStar) ^[3].

A tiny illustration: revenue up $100,000, GOP up $30,000, flow through 30%. The same test works for one department or one offer, using the report's contribution margin after labor: department revenue minus cost of sales and all department labor, including contract labor, divided by department revenue.

Rules of thumb for a good number circulate online without a dataset behind them. Your own trend, year over year, is the benchmark. When revenue keeps climbing and that number does not, it feels like running faster to stand still.

Where is resort revenue growing now?

Rooms are still most of the money. In CBRE Hotels Research's study of 61 US resorts, rooms brought in 56% of 2024 revenue, food and beverage (F&B) 29%, other operated departments 10% and miscellaneous income 5% (LODGING, 2025) ^[6]. Those resorts averaged 280 rooms and a $439.85 average daily rate (ADR), and CBRE says the dataset "favors luxury properties" ^[6].

The growth is outside the room. Across 2,216 US hotels of all types in 2025, CBRE had rooms revenue up 1.1%, F&B up 4.3%, other operated departments up 7.8% and miscellaneous income up 10.9%, against 2.6% for total operating revenue (LODGING, 2026) ^[7].

That fastest line does not prove add-ons are booming. Miscellaneous income is a catch-all, and CBRE credits its growth to "partnerships for unique packages and experiences, to generate commission income and resort/destination fees" ^[7].

CBRE adds that "only resort and all-suite hotels achieved profit growth in 2025" ^[7]. The post on what each staying guest spends beyond the room covers the per-guest metric.

Why did more revenue not mean more profit?

Labor grew faster than revenue. HotStats, a benchmark owned by the revenue management software company Duetto (LODGING, 2025) ^[8], reads profit and loss data from thousands of hotels, sample size undisclosed. It found total revenue per available room (TRevPAR) up 2.7% in 2024 and flow through at negative 4%: GOP fell four cents for every new dollar (2025) ^[2]. Of the $1.04 of extra cost, $0.75 was labor ^[1], so by my arithmetic the other $0.29 was everything else.

The 2025 picture was better, but thin. CBRE had GOP up 1.5% on revenue up 2.6%, the margin slipping from 35.1% to 34.8%, and labor cost per occupied room up 2.9% ^[7]. That is about 20 cents of each new dollar, by my arithmetic from CBRE's rounded figures. CBRE's sample differs from HotStats', so do not chart the two together.

Union status matters. For January to August 2025, HotStats and Duetto found "non-union hotels retain 25 cents in profit, while union hotels lose 1 cent" of each extra dollar ^[9]. Their line for budget season: "achieving even a modest boost in profit margins now requires at least a 5.0% increase in total revenue" ^[9].

When in the year does a resort make its profit?

Seasonal resorts make most of their profit in a short stretch. In CBRE's 61-resort study, summer seasonal resorts earned 59% of annual GOP in their peak months, and winter seasonal resorts 63% (LODGING, 2025) ^[6]. The summer resorts ran 77% occupancy in July and 37.9% in January, and posted "the only negative monthly profit margin in our sample during January 2024" ^[6].

Peak is also when the team has the least slack and extra hours come at overtime rates, so a labor-heavy offer keeps the least exactly when demand is highest. Keep peak offers few and quick to deliver, and test the labor-heavy ones in the shoulder.

Off-season, other operated department revenue held fairly steady, which CBRE reads as a sign that resorts drew local customers to the spa and golf course ^[6]. That is CBRE's inference, not a measured split, so test it against your own point of sale (POS) data first. Miscellaneous income, mostly fixed in nature, was "the least volatile" line all year ^[6].

Why do two benchmarks disagree about the same year?

Owners hear "industry benchmark" as if there were only one. For 2025, CBRE had total operating revenue up 2.6% and the GOP margin down to 34.8% ^[7]. HotelData.com, a vendor platform run by Actabl that does not state its sample, had TRevPAR down 8.8% and revenue per available room (RevPAR) down 6.3% (LODGING, 2026) ^[10], yet a GOP margin up 1.1 points to 38.3%, credited to "labor discipline" (Asian Hospitality, 2026) ^[11]. For 2024, CoStar had TRevPAR up 7.2% (2025) ^[12] and HotStats 2.7% ^[2].

None of them is wrong. They measure different hotels, and sometimes different things: total revenue is not revenue per available room. The mistake is mixing them, setting your year against one source's last year and another's this year.

Four questions before you quote a benchmark to an owner:

1. Whose hotels are in the sample: type, size, union or non-union, region?
2. Which measure is it: total revenue, TRevPAR, RevPAR or GOP margin?
3. Is it the same source you used last year?
4. Is the publisher independent or a vendor, and does it state its sample?

Which offers flow through best?

Most likely early check-in and late checkout, made from rooms and a housekeeping plan the resort already has: no cost of sales, little labor.

They are also what guests buy. In Akia's vendor data on 4,729 paid upsells, pooled across hotels and vacation rentals, the two made up 90% of upsell revenue and the most common price paid was $50 (2026) ^[13]. The data has no F&B, spa or parking items. No public source measures profit by offer, so this is an inference from cost and sales data, not a finding.

Volume is low, so price and timing decide the dollars: 1.59% of stays at properties running upsells bought one, 3.66% for guests booking 31 to 60 days ahead and 0.89% for same-day arrivals ^[13]. Akia also found 42% of early check-ins bought on arrival day and 49% of late checkouts on departure day (2026, sample not stated) ^[14]. Pricing late checkout by the hour gets its own post.

Tone matters too: the moments guests say yes to an offer are rarely the ones a busy desk finds convenient.

What I resent as a guest is the folio line I never agreed to. Since May 12, 2025, the Federal Trade Commission's fee rule has required lodging businesses to show the total price more prominently than any other price; optional extras may be excluded [15]^[16]. A mandatory fee is part of the price now, and an optional extra has to earn its yes. This is general information, not legal advice.

What does flow through look like for two offers, side by side?

The dollar test, for any offer:

1. Start with the price.
2. Minus the cost of what is consumed.
3. Minus labor hours at the loaded wage, at overtime rates in peak.
4. Minus card fees and supplies.
5. Divide what is left by the price.

In this illustration, the late checkout keeps $40.50, or 81%. The beach dinner keeps $56.80, or 24%, in a normal month, close to the 29.1% F&B department margin CBRE measured for the first half of 2025, when labor was 59.4% of department expenses ^[17]. So the assumptions are not rigged against F&B. When its four labor hours come from peak overtime, it keeps $8.80, or 4%. Only the $50 is sourced: it was the most common price paid in Akia's 2026 vendor data, not a recommendation ^[13].

Dinners are not bad. They are labor-heavy and season-sensitive, and profit per labor hour shows it: $121.50, $14.20 and $2.20.

Now the volume. A hypothetical 120 room resort at 70% occupancy and a 3.5-night average stay has about 8,760 stays a year. At Akia's pooled 1.59% (2026) ^[13], that is about 140 upsells, nearly all time offers; at the early bookers' 3.66%, about 320. Useful money, not a rescue plan. Then run your own assumptions in the ROI calculator.

Why is department profit, not revenue, the scorecard?

Because departments keep very different shares of a dollar. In CBRE's 2023 data on 1,178 hotels, parking department profit was 61.3% of parking revenue, while "the average profit margin for all other-operated departments was 58.7%" (2024) ^[18]. The F&B department margin in CBRE's data was 29.1% in the first half of 2025 ^[17]. Different samples and years, so read them as rough levels.

The spa shows the trap. Across 297 US hotels that run their own spa, CBRE found 2024 spa revenue up 1.4%, department expenses up 2.1% and spa labor up 3.9%, while spa department profit fell 0.5% (Hospitality Net, 2025) ^[19]. Revenue up, profit down.

Put department profit after labor beside revenue in the monthly owner report, split into peak and off-peak months. For the rest of the agenda, see the numbers that survive an owner meeting.

To start, run the two-hour ancillary scorecard with your own property management system (PMS), POS and payroll reports.

Where ResortConcierge AI helps

The practice is a posted price, sent when guests decide, and confirmed by someone who can see the room plan.

ResortConcierge journeys send timed messages before arrival, at check-in, mid-stay and at checkout, so a posted early check-in or late checkout price can reach guests in the final week and again on the day. It needs a connected email, text or WhatsApp channel to send.

The answer should not depend on who picks up. When a guest asks in the conversation, the AI concierge answers from the price and rules the property has written into its knowledge base, so the answer is the same at 7am and 11pm, and a person still confirms against the room plan. It is part of running a resort on margin, not volume.

PMS connectors are planned. Until then, posting the fee to the folio and checking room status are staff steps.

More revenue only helps when it costs less than it brings in, and at US hotels in 2024 it did not. I will keep paying for the 4pm checkout; I would just like the resort to keep most of it. Pick one extra you sell today, write down its price, its cost and the hours it takes, and do the dollar test before peak.