Comparing This Year's ADR With Last Year's After a Fee-Structure Change

Two years, one chart, and a fee base that moved underneath both of them. The comparison is still possible — it just needs three adjustments first.

Last updated: September 22, 2026

Average daily rate is a division: booked revenue over nights sold. What changed underneath it, on one of the four channels a host typically sells on, is what the guest pays and what reaches your account. Under the Airbnb announcement of 7 July 2026, the previous split structure became a single 15.5% service fee borne by the host, and unlike any previous version it applies to the whole booking subtotal — cleaning fee, pet fee and other host-charged extras included. Everything below that line shifted with it. This article sets out which reported number actually moved, the three adjustments that make two years speak to each other again, and a one-page comparison you can rebuild every quarter without re-learning how you did it last time.

Key Takeaways

• The base widened, not just the rate. The fee now reaches the cleaning fee and other extras, so comparable money sits in a different place.
• ADR did not break; the definition drifted. Same division, different input, once the fee moved.
• Three adjustments, in order. Normalise the base, choose one rate basis, then match the night mix.
• Occupancy can rise while comparable rate falls. Shorter stays and discount-led booking do that without any pricing error.
• Rebuild the page quarterly. A comparison you can reproduce beats a number you have to argue for.

What actually moved, and which number on the report it moved

Say the mechanism plainly, because almost all of the confusion downstream comes from it. Under the Airbnb announcement of 7 July 2026, the former arrangement of a host-side charge alongside a guest-side service fee became a single 15.5% service fee borne by the host, with the guest no longer seeing a separate service-fee line at checkout. Brazil and Mexico stayed at 16% under the same announcement. There was no opt-out, and the migration ran to different dates in different places: 15 September 2026 outside the EEA and 13 October 2026 in the EEA and Switzerland. Those dates and rates are exactly the kind of thing to confirm against what your own dashboard currently shows rather than any article, including this one.

The part that has consequences for your reporting is the fee base. It is the whole booking subtotal, cleaning fee and other host-charged extras included, with taxes and deposits outside it. That has two effects. First, a stay with a large cleaning component now carries more fee than the same night did before, which changes the comparison between properties and between seasons. Second, and more confusingly, many hosts repriced at the same time to protect what they took home — and the arithmetic behind that repricing is not what people assume. Under the announcement, keeping the same net requires the old list price to rise by roughly 14.8%, or equivalently dividing your target take-home by 0.845, about an 18.3% gross-up. Simply adding 15.5% to the old rate lands in neither place. Whether you repriced correctly is now baked into your revenue and therefore into your ADR.

What this means practically is that the number that moved is not the reported nightly amount. It is everything downstream of the channel statement: what you considered the rate, how much of the guest's payment reached you, and which comparison you were making when you read that line. Availability and rates for Airbnb, Booking.com, Agoda and Trip.com sit on one grid at localsbnb.com, which matters here because the repricing has to be applied to identical nights across all of them or the comparison acquires a third moving part.

The three adjustments that make two years comparable

Do them in this order. Doing them in a different order produces a plausible-looking number that nobody can defend.

Adjustment	What it fixes	How it is done	What breaks if skipped
Normalise the fee base	Cleaning fees and extras now sit inside what is charged on	Restate both years with the same treatment of extras	Season-to-season swings that are really composition
Choose one rate basis	One year reads net, the other reads gross	Pick net-of-fee or gross and rebuild both years on it	A jump or drop that is purely definitional
Match the night mix	ADR divides total by nights	Compare the same natural weeks, units and stay lengths	A pricing problem that is really a length-of-stay shift

The first adjustment deserves a warning about effort. Doing it properly means restating last year, which is tedious. Doing it approximately is worse than not doing it, because an approximate restatement leaves a small unexplained difference you will spend next year chasing. One afternoon with last year's statements is enough.

The second is a decision rather than a calculation, made once and written down. Either every figure you compare is net of that channel's fee, or every figure is gross. Write whichever you chose at the top of the page, because in six months neither you nor a colleague will remember. The home dashboard reports occupancy, ADR and RevPAR together, and those three stay consistent only if the ADR feeding them is defined once.

The third is where most year-on-year mistakes actually live. Benchmark reports published in September 2026 — including PriceLabs' September benchmarking notes, which are an industry source rather than a platform rule — compare properties on their own definitions, and those rarely match a single host's unit mix. If one side of your comparison contains more units, or a different blend of one-night and week-long stays, the average moves without anybody changing a price.

Why occupancy can rise while comparable rate falls

This is the pattern that makes hosts distrust their own reporting, and it usually has nothing to do with the fee change.

Occupancy counts nights sold. ADR averages revenue over those nights. Fill a property with shorter stays and you raise the first while lowering the second, because fixed costs of a changeover — cleaning, laundry, the coordination — are spread over fewer nights of revenue. It is entirely possible to work harder, sell more nights, be busier in every way your calendar can show, and be measurably worse off. Discounts have the same effect with less effort: they buy occupancy at a rate you would have preferred not to grant.

The other cause is mix rather than behaviour. A property that once attracted week-long bookings now fills with two-night stays: occupancy improves, ADR falls, and if those bookings arrive through a different channel the effect compounds. Before concluding that your pricing is wrong, check whether length of stay moved. It is a one-line query and it explains more of these gaps than any repricing story does.

The one-page comparison you rebuild each quarter

Keep it to five lines and one note, and rebuild it the same day every quarter so the previous build is still fresh enough to imitate.

Five lines: nights sold this natural quarter against the same natural quarter last year; booked revenue on the gross basis and the net-of-fee basis; the resulting ADR under each; occupancy for the same two quarters; RevPAR, which is where the two squared off against each other. Then one note — anything structural worth remembering, such as a unit added, a channel opened, or a week taken off sale for repairs. Comparing columns that include a week you deliberately did not sell is how good decisions get described as bad ones later.

RevPAR is worth reading third rather than first. It is the line where a fall in ADR coexists with improvement, and where rising occupancy is revealed as discount rather than demand. Occupancy, ADR and RevPAR sit together on the home dashboard because reading one alone invites the wrong conclusion. If you want those three measured on a definition you wrote down, rather than one you have to reconstruct, that is something to look at first-hand at localsbnb.com.

FAQ

Do I add 15.5% to my old rates to keep the same net?

No. Under the Airbnb announcement of 7 July 2026, protecting the same take-home needs roughly a 14.8% rise on the old price, or dividing the target net by 0.845 — about an 18.3% gross-up. Adding the fee percentage directly lands between the two and matches neither. Confirm current figures and dates against what your dashboard shows.

My ADR fell but I did not change my pricing. What should I check first?

Length of stay, then unit mix, then channel mix. ADR divides revenue by nights, so more one-night bookings lower it without any rate decision at all. Only after those three come back level should you treat it as a pricing problem.

Should I use gross or net-of-fee figures for my own comparison?

Whichever you will stick to for a year. Net-of-fee answers "what reached me" and is usually the honest line for judging performance; gross answers "what did the guest pay" and is better for positioning. Pick one, write it at the top, and rebuild both years on it.

A year-on-year comparison is not supposed to be comfortable; it is supposed to be repeatable. Do the three adjustments once, write down the basis you used, and rebuild the page on the same day each quarter. The next time someone asks why this year looks worse, the answer is a page rather than a guess. If you want occupancy, ADR and RevPAR living on one definition instead of four spreadsheets, you can start free with LOCALSBNB.

---

Figures, dates and migration schedules above are stated per the Airbnb announcement of 7 July 2026 and should be confirmed against the current notice in your own host dashboard, which governs. Benchmark references are industry sources. This page is information, not financial or legal advice. LOCALSBNB provides software, not legal advice.