Two Airbnb checks, one quarter to earn them
Situation. Airbnb is paying twice this quarter. It pays $15–$20 for every listing we enroll in the Highly Rated Guest discount by Dec 31, with the $20 rate unlocking once 10% of our eligible listings are enrolled. Separately, it pays for extra Q4 nights booked on a named set of 14,700 of our Airbnb listings, with no earnings cap this year.
Problem. Enrollment is small today (393 accounts, an estimated 2,500–3,000 listings) and the 14,700-listing cohort is set up for summer, not Q4: the median listing has 58 of its next 77 nights open, 69% have no weekly-or-longer discount, and 23,439 nights are stranded in 1–3 night gaps between bookings.
Answer. Win in three moves. One, convert the hosts who already said yes to the same idea on another channel, 1,232 accounts run a Vrbo audience discount with no Airbnb equivalent, using one-click enrollment with their peak nights pre-protected. Two, unlock stranded nights: turn on the gap rule we already ship, add gap-night pricing, push monthly discounts in Florida, and prompt owners to release stale blocks. Three, ship a pricing cockpit (pacing plus unbookable inventory, every channel and direct), so hosts see their Q4 building week by week and stay enrolled through Dec 31.
Where to go: Deciding? Read The biggest levers and Decisions needed. Selling or supporting hosts? Read How an expert runs this and What changes for each role. Building? Read How it works. Checking the numbers? Read What we found and Data and method.
The problem: the money is real, the clock is short, and the host pays
Eleven weeks remain, a third of the pilot window is already gone, and neither incentive pays for good intentions.
The Highly Rated Guest pilot runs Sep 14–Dec 31, 2026. Listings enrolled before we cross 10% earn $15 each; every listing enrolled after crossing earns $20, provided we are still at 10% or above on Dec 31. Listings enrolled before the pilot earn nothing but count toward the gate. Our Airbnb base is roughly 41,000–49,000 listings, so the gate sits near 4,100–4,900 enrolled listings. Today 393 accounts run 454 Highly Rated Guest discounts, covering an estimated 2,500–3,000 listings. The gap to the gate is plausibly 1,500–2,000 listings, close, but not close enough to leave to organic growth.
The Productivity incentive names 14,700 already-booked Airbnb listings (the cohort locked Jun 20, 2026) and pays for incremental nights booked on them in Q4, now with tiers above last year’s cap. Those listings booked 303,527 Airbnb nights in Q4 2025. They hold 145,305 Airbnb nights for the rest of Q4 2026, with most of the booking curve still ahead. Every extra point of occupancy on this set is roughly 11,000 nights.
One positioning rule sits under everything host-facing: we help these listings book more overall, across Airbnb, Vrbo, Booking.com, and direct. Airbnb then wins its share of a bigger pie. Every host-facing tool below (pacing, gap pricing, discount wizard, rate compare) is channel-agnostic on purpose; the incentives simply pay us to build and promote them now.
One tension sits under the first incentive: we earn $15–$20 and the host gives 15% off. A pitch that hides that will fail with the first sharp property manager. Enrollment has to be a good deal for the host on its own terms, better guests, peak nights protected, more bookings, with a share of the bounty as the sweetener, as we have shared before. The good news: this week we shipped the feature that answers the strongest objection. Blackout dates (the Stay Date condition) let a host exclude up to 30 nights a year per listing from the discount. “Not on Christmas week” is now a setting, not a refusal.
The proposal in one paragraph
Run Q4 as a named-account campaign, not a broadcast. In week one, cut five target lists we can already produce from production data, Vrbo-discount cross-sell, deactivated enrollments, Airbnb Mobile cousins, the top 200 managers by cohort listings, and Last-Minute/Early-Bird hosts with no monthly discount, and start the zero-build campaigns (gap-rule turn-on, block release, monthly ladder for Florida). In weeks two to four, ship the conversion layer: one-click mirror from a Vrbo discount to its Airbnb twin, one-click reactivation, and enroll-all with each listing’s priciest nights pre-checked as blackouts (annual slider, current year prorated). In parallel, scope the two real builds, gap-night pricing and a multi-channel rate compare, and stand up the pricing cockpit that reports bookings pace to hosts and enrollment pace to us. Sequence the big manager payloads just after we cross 10%, when each listing pays $20 instead of $15, and defend the gate through Dec 31 with a disenrollment save play.
Words we use
- Highly Rated Guest discount
- Airbnb’s promotion, called Top-Rated Guest Discount in the incentive contract. A fixed 15% off, set by Airbnb, shown to guests with strong track records. In OwnerRez it is a discount type a host adds per listing set.
- Blackout dates / Stay Date
- Nights a discount does not apply. Our Stay Date condition (“Stay Date anything but …”) sends them to Airbnb (max 30 per listing per year) and to Vrbo. Shipped this week on all four audience discounts: Vrbo Mobile, Vrbo OneKey, Airbnb Highly Rated Guest, and Airbnb Mobile.
- 10% gate
- The contract threshold: once 10% of our eligible Airbnb listings are enrolled, new enrollments pay $20 instead of $15, if we are still at 10% or more on Dec 31.
- Productivity cohort
- The 14,700 Airbnb listing IDs Airbnb named for the Q4 nights incentive (their sheet, Oct 6). We matched 14,693 to OwnerRez properties: 14,688 properties, 3,435 accounts.
- Gap night (orphan)
- An open run of nights trapped between two bookings or blocks, too short for the property’s usual minimum stay.
- Min-gap rule
- Existing OwnerRez setting: inside a gap, the minimum stay relaxes to the gap length so the nights can book. Only 23% of cohort listings have it set.
- Spot rate
- A per-night price (and optionally min-stay) written for specific dates, by the host or by a pricing tool through our API. Spot values override seasonal and property defaults for that night.
- Rate adjuster
- A per-channel percent markup or markdown on base rates (a surcharge type in our data). How hosts offset channel commission differences.
- Impact Data
- Airbnb’s open ask to show hosts their results data (bookings, nights, program effect) inside OwnerRez. Paul Hall owns the ETA. Our answer: the generic pricing cockpit, filtered to Airbnb where needed.
- Pacing
- How much of a future stay window is booked by a given date, versus the same point last year. The scoreboard hosts actually want, across all channels and direct.
- Bounty share
- Passing part of Airbnb’s per-listing enrollment payment to the host as an OwnerRez credit. We have shared incentive bounty with hosts before.
What we found in the 14,700
The cohort is concentrated three ways, by place, by manager, and by season, and each concentration is a targeting gift.
This is a Florida and Smokies cohort, not a national one
Listings per state, 14,688 matched properties; 91.2% are in the US. Top metros: Panama City Beach 953, Sevierville 752, Kissimmee 475, Gulf Shores 410, Myrtle Beach 408. Q4 means snowbirds in Florida and holidays in the Smokies, two different offers.
Two hundred accounts hold 45% of the cohort
Cumulative cohort listings held by the largest accounts. The largest single account holds 118 cohort listings; 330 accounts hold 10 or more (56% combined). A named-account motion reaches half the cohort through a few dozen conversations.
The nights are there; the setup is not. For Oct 16–Dec 31 the cohort holds 774K open nights (median 58 open of 77 per listing) against 254K booked across all channels and 103K under owner blocks. Blocks already exceed all of Q4 2025. Inside the booked calendar sit stranded gaps:
63K nights are stranded in gaps of six nights or fewer
Bounded gaps (booking or block on both sides), Oct 16–Dec 31, 2026. The 1–3 night pool alone is 23,439 nights on 39.5% of listings, and it grows with every new booking. In 41.3% of those gaps the minimum stay exceeds the gap, so the nights cannot book unless the min-gap rule is on. It is on for 23% of listings. The calendar warehouse separately flags 7,811 available nights where no valid stay exists at all under current rules: the truly stranded subset, and a ready-made targeting flag.
Most of the cohort has no length-of-stay ladder at all
Share of 14,688 cohort listings covered by an active Airbnb-applicable discount of each kind. One 28-night booking is 28 incentive nights; Florida snowbird demand books in the next six weeks.
Enrollment headroom is warm, not cold
- Cross-sell is the whale. 1,232 accounts run Vrbo Mobile or OneKey (Vrbo’s member program) with no Highly Rated Guest discount. They already accepted the audience-discount idea; 639 of them are also in the Productivity cohort. The reverse (Highly Rated Guest, no Vrbo promo) is 188 accounts. Airbnb Mobile without Highly Rated Guest: 163 accounts, the closest cousins of all.
- Resurrection pool: 194 deactivated Highly Rated Guest discounts across 166 accounts (live count; it moved by one during this analysis). In the cohort alone they would cover 317 unenrolled listings. These hosts said yes before blackout dates existed.
- Only 6.8% of cohort listings are enrolled today (992 listings, 176 accounts), so the overlap play (enroll cohort listings first, earn bounty plus nights) is almost all headroom.
- Enrolled listings book more, with a caveat. Enrolled cohort listings averaged 23.9 Airbnb nights in Q4 2025 vs 20.4 for the rest. Enrolled accounts skew larger and more sophisticated, so treat this as directional, never as a promise to a manager.
Pre-filled blackouts have a real signal
93.2% of cohort listings carry spot rates for the next 365 days (71% of those rows are written by pricing APIs), and 99.3% have spot or seasonal rates. On a 295-calendar sample, a listing’s 30 priciest nights hold a median 16.0% of its annual rate value, double the 8.2% that flat pricing would give, and the cheapest of those 30 still prices at 1.9× the median night. Resolving each listing’s own forward calendar and pre-checking its top 30 nights gives the host a suggestion they recognize on sight: their own peak weeks, protected in one click.
Friction, ranked by size
- Minimum stay of 3+ nights: 39.3% of listings, heading into a short-trip season. 7+ nights: 7.0%.
- Owner blocks: 103,068 nights in the remaining window, on 32.8% of listings, plus 85,203 stood nights already looming over Jan–Mar 2027.
- Min-gap rule off: 77% of listings cannot take a short stay inside a gap.
- Cancellation and Instant Book, resolved through channel defaults (listing override, else the connected account default; 94.5% matched): Firm 45.1%, Strict-14 21.7%, Firm-14/Limited 11.2%, Moderate 14.0%, Flexible 4.0%. Instant Book: Everyone 60.3%, Well-Reviewed 28.8%, Request-to-Book 967 listings (6.6%). Seasonal cancellation policies (shipped for Airbnb and Vrbo) run on only 625 listings (4.3%, 127 accounts) and today skew stricter-in-season. The unused direction, strict at peak and relaxed in shoulder, is the lever: hosts who will never go Flexible year-round will often go Moderate for the eight softest weeks.
- Non-issues in this cohort: prep days (119 listings), long lead time (~220), arrival-day restrictions (~4%).
Listing quality is a tail problem: and a good door-opener
The median cohort listing has 42 photos at 2255px median width, and 57% are Airbnb Guest Favorites. The fixable tail: 3,085 listings (21.0%) have zero captioned photos; 1,164 (7.9%) set no location/view flags at all; 500 (3.4%) carry three or more near-duplicate photos; 294 (2.0%) have mostly sub-1024px photos; bedrooms are unset on 3.1%. A per-manager fix list built from these rules is a genuine give that opens the enrollment conversation. Photo-vs-amenity inference (does the photo show a view the listing never claimed?) is a sound second phase.
Airbnb is rarely the expensive channel by adjuster: except where it is
Among cohort accounts with both an Airbnb and a Booking.com adjuster (117 accounts), 78.6% set them exactly equal; the median difference is zero. Fifteen accounts differ by more than 5 points. The cell worth acting on: 144 accounts run an Airbnb-only adjuster at a median +15% with no Booking.com adjuster, if those listings sell on Booking.com at base rate, Airbnb is systematically their priciest channel. That is the shape of the price-competitiveness complaint, and it is fixable account by account. With Booking.com commission near 18% (program-dependent) against Airbnb’s 15.5% host-only fee, net parity puts Booking.com roughly 3 points above Airbnb; a ±5% all-in band is a fair warning line. Adjusters alone will not catch tax and fee drift, which is why the compare tool below matters more than a warning.
Pacing: December books late, and this year booked early
Guests book this cohort close in: median lead time is 23 days on Airbnb (19 for Q4 arrivals), 24 on Booking.com, and 43 on Vrbo. Bookings made after Oct 15 last year arrived a median 7 days later. That is the working window for everything below: changes live by early November catch most of the remaining demand; December changes catch only the last-minute tail. Scope note: every pacing figure in this section is the 14,688-listing incentive cohort, not the platform.
Gross booking value (GBV, what guests booked in dollars) tells the same story louder. Same-store (listings with Q4 2025 bookings), Airbnb GBV on the books at Oct 15 is $50.3M vs $36.5M at the same point last year (+38%), on +20% nights, with on-books average nightly value (ADR) $445 vs $386 (+15%; last year’s final ADR diluted to $333 as cheaper late nights booked). All-channel plus direct, same-store GBV is +37%. Market sentiment can be soft while this selected cohort runs hot; both can be true, and the warehouse’s independent booked-night count corroborates the Booking-derived curve.
Most of Q4 books after Oct 15, and December barely books before it
Share of final Q4 2025 Airbnb nights (surviving bookings, 252,740 nights) that were on the books by Oct 15, 2025, by stay month. December is the quarter: 77% of its nights booked after Oct 15.
This year is pacing ahead and earlier: 145,305 Airbnb nights on the books vs roughly 107K at the same point last year, with the booked-by-Oct-15 cohort up 28% in count and its median lead stretched from 66 to 73 days. Repeating last year’s curve implies a 344–359K final (+40%); repeating last year’s final implies 108K nights still to book. The front-loaded curve and the 17.3% of listings with no Q4 2025 bookings (new or dormant then) both say the curve method overstates. Plan on a wide honest range: 110–210K Airbnb nights still to book naturally, most of them for November and December stays. Natural pickup supplies most of the remaining nights this quarter; our levers compete for share of a booking wave that is already forming, which is exactly when conversion fixes pay best.
Blocks mostly open up, and they open late
Hosts do block the season and release it. Blocks touching last Q4 totaled about 208K nights; 59% of those block-nights were released before stay time, and 47% of freed nights rebooked. The waste is timing: releases land a median 20 days before arrival, 38% inside the final week, and 8% after arrival (retro cleanup). A night freed five days out mostly misses a 23-day-median booking curve. New blocks also keep arriving all quarter (creation lead median 83 days, 17% created inside 30 days), so the pool still standing (103K nights today) will shrink by gross releases, not collapse. Composition says mostly deliberate: blocks that stood (were never released) are dominated by short owner-style holds (median 3 nights last year) plus 31+ night seasonal blocks (67% of currently stood nights, 1,304 listings), 63% of this year’s blockers blocked last year too, and the top recorded reason is blocks imported from Airbnb itself. The mistake-shaped residue is small and listable: last Christmas week, 1,078 strong-booking listings stood blocked for 3,972 nights, the largest reason bucket blank. That is an audit-and-ask list for CS, not a campaign.
Imported blocks deserve their own read. 48,550 block rows carry an import reason (top exact text: “imported owner block from airbnb”), concentrated on 1,902 listings at a median 5 per listing. They are bimodal: median length 2 nights, yet 20% run 90+ nights, with a median creation-to-arrival lead of 365 days. That shape fits rolling 12-month channel closures synced in from Airbnb and other sources more than it fits a max-window artifact (only 17% arrive within three days of created-plus-booking-window). They churn hard (only 19% still stood vs 48% of non-imported) and are just 4% of currently stood Q4 nights, so for this quarter they are mostly self-healing system noise. The looming unbookable pool is not imported: Jan–Mar 2027 already carries 85,203 stood nights, 70,833 of them in non-imported 31+ night blocks. Accidental versus purposeful cannot be read off one field, but three classes separate cleanly enough to act on: imported rolling closures (leave, watch), repeat seasonal mega-blocks (deliberate; ask once when their dates enter the 45-day booking window), and the suspicious residue (blank-reason prime-week blocks on strong bookers, blocks released only after arrival). The cockpit below exists to work the second and third classes by name.
Two-thirds of the cohort is dynamic-priced, and it is mostly PriceLabs
9,569 listings (65.1%) have an API app as their dominant spot-rate writer over the next 365 days: PriceLabs on 8,638 listings, Beyond on about 518, Wheelhouse on 254, then a long tail. APIs write 71% of all spot rows. Dynamic-priced listings carry fewer warehouse-orphan nights per listing than manual ones (0.45 vs 0.69) at the same 18% incidence: pricing tools soften gaps but cannot fix min-stay rules, which live here. Control note: tools write MinNights on 91.7% of their spot rows, and spot min-stay wins nightly precedence over property and season rules, so for 61.6% of the cohort the pricing tool effectively is the min-stay policy. The saving grace for gap work: tools set the suppress-gap-relaxation flag on only 1.31% of rows, so our min-gap relaxation usually survives tool pricing. Two consequences. First, gap features must layer visibly over tool-written rates or hosts will not trust the stack. Second, PriceLabs is a distribution partner hiding in plain sight: 8,638 shared listings is co-marketing scale.
Cross-sell accounts are already on both channels
Channel presence, measured from live channel connections and last-12-month bookings, says the mirror pitch rarely needs a channel sale first. Of the 1,232 Vrbo-promo accounts without the Airbnb discount, 91.4% are Airbnb-connected (8,526 active listings) and 93.5% took Airbnb bookings in the last year; only 106 have no Airbnb at all. The 163 Airbnb Mobile cousins are 98.8% connected. The reverse direction is thinner, as expected: of 188 Highly Rated Guest accounts with no Vrbo promotion, only 58.5% are Vrbo-connected. Segment files with per-account listing counts accompany this page’s data pack.
The biggest levers, ranked
For enrollment, sell to the convinced first and cross the $20 gate early. For nights, unlock inventory that is already stranded before chasing new demand.
Yields marked est. carry the stated assumption. Segments overlap, do not add the rows.
Enrollment ladder: pays $15, then $20, per listing
| Lever | Audience | Expected yield | Effort |
|---|---|---|---|
| 1. One-click mirror: Vrbo discount → Airbnb twin. In-app button at discount save and on the discounts grid, announced with “blackout dates on all four promotions.” | 1,232 accounts | Measured: 1,126 of them are already Airbnb-connected, holding 8,526 listing-active Airbnb listings (median 3 per account). If a quarter enroll most of theirs: est. ~2,100 listings. Likely crosses the 10% gate with help from rows 3 and 4. | Small build |
| 2. Enroll-all with pre-filled blackouts. Every segment converts through this door: priciest nights pre-checked on a prorated annual slider, time-boxed “on through Dec 31” framing, bounty-share amount shown before the click. | All segments | Multiplier, not a segment. It is the difference between a settings project and a ten-second yes. | Small–medium build |
| 3. Manager packets for the top 200. Named humans, per-manager numbers: their listings, open Q4 nights, protected peak nights, bounty-share dollars, quality fix list. | 200 accounts / 6,644 cohort listings | Top-50 accounts average ~60 cohort listings each. Est. 1,500+ listings if half of the top 50 enroll most of their portfolio. Sequence just after the gate so these pay $20. | CS/sales time, no build |
| 4. Resurrect the deactivated. “The reason you switched it off is fixed” + one-click reactivate with suggested blackouts. | 166 accounts | Measured: 1,594 listing-active Airbnb listings sit behind 161 of those accounts (317 of them in the Productivity cohort). At half returning: est. ~800 listings, the cheapest enrollments we will get. | Tiny build |
| 5. Default-on at Airbnb connect and listing import (suggested blackouts attached, clearly opt-out). | New listings, Oct–Dec | Compounding tail; every new listing starts enrolled while the pilot pays. | Small build |
| 6. One-button Airbnb launch for Vrbo-only accounts. “You are on Vrbo; fire up Airbnb from the listings you already have here,” with content, rates, and discounts carried over and enrollment on at birth. | 106 accounts / 1,180 Vrbo listings in the cross-sell segment alone | Slower per account (connect plus publish, then enroll), and the timing is a gift: hosts are rethinking channel mix after Vrbo raised its host fee to 12%. Every converted account also feeds the Productivity side next year. | Connect-flow assist + campaign |
Illustrative prize math: at 8,000 total enrollments with the gate at ~4,500, payout is roughly 4,500 × $15 + 3,500 × $20 ≈ $138K. At 12,000:≈ $218K. Bounty share and campaign costs come out of that, which is the point: acquisition spend per listing has a hard ceiling of $15, then $20.
Nights ladder: pays per incremental Q4 night on 14,700 listings
| Lever | Pool | Expected yield | Effort |
|---|---|---|---|
| 1. Gap machine. Campaign to switch on the existing min-gap rule (77% have it off), then ship gap-night pricing so stranded nights are priced to move. | 23,439 nights (1–3n) + 39,534 (4–6n) | At 10–20% conversion of 1–3n gaps: est. 2,300–4,700 nights in this window; the pool refills with each booking. | Campaign now; medium build |
| 2. Monthly ladder, Florida first. One-click 7/14/28 tiers, pitched to the 601 accounts that run Last Minute or Early Bird with no monthly tier, then to all 5,146 Florida listings before snowbird booking peaks. | 2,537 warm listings; 10,182 with no length-of-stay tier at 7+ nights | If 5% of the warm segment books one 28-night stay: est. ~3,500 nights. Value doubles if the incentive counts booking date (open question). | Template + campaign |
| 3. Use-it-or-release-it block prompts. Surface each listing’s blocked Q4 nights and ask, 30 days out, whether they are still needed. | 103,068 blocked nights | 5–10% released and half rebooked: est. 2,500–5,000 nights. | Email + in-app prompt |
| 4. Pricing cockpit: pacing plus unbookable inventory, all channels plus direct. “Nights booked for Q4 vs the same point last year” per host and per manager, with channel mix and gap and block alerts, plus our internal pace ticker. Deliberately not an Airbnb scoreboard: hosts manage a whole business, and the direct-booking crowd must see itself in it. Airbnb gets the same data sliced to its listings, which answers the Impact Data ask without building an Airbnb-only surface. | All 3,435 cohort accounts | Force multiplier: keeps hosts enrolled to Dec 31 (protects the $20 tier), makes every other lever visible, and outlives the incentives as a core reporting surface. | Small build |
| 5. Friction fixes by list, not by article. Min-stay 3+ in short-trip markets, explicit Request-to-Book, strictest cancellation, each host gets their own numbers in the packet, not a generic checklist. Includes seasonal cancellation (strict peak, relaxed shoulder) for the 95.7% not using it, and booking-window extensions where the window is shorter than the market’s typical lead time. | 5,769 min-stay-3+ listings | Diffuse but real; rides free inside the manager packets. | Content + data pull |
Scale check: the cohort booked 303,527 Airbnb nights in Q4 2025. The stack above is a plausible 3–5% incremental lift (10K–15K nights), before any price-competitiveness work. The two ladders multiply on the same listings: an enrolled cohort listing earns the bounty once and better odds on nights all quarter, which is why cohort listings go first in every enrollment segment (639 cross-sell accounts are in both sets).
How an expert runs this
Segment by intent, make the yes reversible, and treat Airbnb as a distribution partner, not just a payer.
- Sell identity, then mechanics. “Top-rated guests for top-rated homes” speaks to the 57% who are Guest Favorites and the 1,148 listings already set to well-reviewed-guests-only Instant Book. The 15% is the price of a better guest mix, fewer parties, fewer damage claims, and blackout dates cap the downside. Lead with guest quality; put the arithmetic one click deeper.
- Every ask is one click, reversible, and time-boxed. “On through Dec 31, off automatically in January unless you keep it” beats a permanent settings change. Pre-filled blackouts mean the host reviews instead of creates.
- Land and expand inside each manager. Enroll five listings, report guest quality and bookings after 30 days from their own data, then ask for the portfolio. Managers also need an owner-facing one-pager: many cannot discount owner revenue on their own signature. Forgetting the owner is the classic way this motion stalls.
- Sequence around the gate. Sprint the self serve segments (mirror, resurrect) to cross 10% fast, then land the manager payloads at $20. Publish the countdown internally; a visible gate focuses a quarter like nothing else.
- Cap acquisition spend by the bounty. Each enrollment is worth $15, then $20, minus the host share. That sets the price of doing it by hand. An hour of human time needs to enroll roughly a handful of listings pre-gate, fewer after.
- Trade roadmap for distribution. Airbnb wants adoption and price competitiveness. Offer Dianna Macwan, our Airbnb partner lead, a dated plan, blackouts shipped, mirror + enroll-all in November, Impact Data in the UI (Airbnb’s open ask to show hosts their results data inside OwnerRez; Paul owns the ETA), multi-channel compare scoped, and ask in return for host-facing placement (their email and in-product reach to our hosts), search-impact proof we can quote, and the tier table so we can set public internal goals. The checks end Dec 31; preferred-partner standing should not.
- Equal opportunity, always. We align campaigns with partners, and we stay the platform that helps hosts book everywhere, direct included. Frame every pushed night as incremental (a gap night, a shoulder weeknight, a snowbird month that would otherwise sit empty), never as moving bookings between channels. The moment a host hears “move bookings to Airbnb,” we lose the room and the account. Tools that only make sense for Airbnb (enrollment, blackouts) are the exception; pacing, gap pricing, discounts, and rate compare must work for every channel.
- Retain to the measurement date. The $20 tier survives only if we hold 10% on Dec 31. Weekly pacing reports, a disenrollment alert with a same-day save attempt, and a mid-December “you are N nights from last year’s Q4” nudge protect the whole tier, not just one listing.
What the data knocked loose
Four ideas were not in the first brainstorm; the numbers insist on them.
- Two geo kits instead of one national blast. A Florida Snowbird Kit (monthly ladder, long-stay cancellation check, block release) for 5,146 listings, and a Smokies Holiday Kit (2-night minimums, gap fill, Thanksgiving-to-New-Year pricing) for the Tennessee cluster. Same machinery, local copy, local proof.
- Caption sprint as the value-give. 3,085 listings have no photo captions. Draft captions for approval in bulk (rules first, vision model where it earns it), bundle with the view-flag and duplicate-photo fixes, and hand each manager a finished fix list for their own listings. Generosity opens doors that pitches do not.
- “Beat your Q4 2025” as the host-facing frame. We now hold the per-listing baseline. A weekly “you are at 12 of last year’s 22 Airbnb nights, with 3 gap nights open” message turns an abstract incentive into a game the host is already playing.
- A parity rescue list for Dianna. The 144 Airbnb-only-adjuster accounts are a named, fixable answer to the price-competitiveness ask. Fix them with the hosts, then show Airbnb the before/after. Few partner gifts beat a closed complaint with receipts.
- PriceLabs is a channel, not just a vendor. 8,638 cohort listings price through PriceLabs. A joint webinar and email on gap nights and pacing reaches two-thirds of the cohort through a tool they already trust, and the trade is honest: their orphan pricing works better when our min-gap rule lets those nights book at all.
- A stranded-nights inbox. The warehouse already flags orphan nights nightly. Surface them where hosts work (“3 nights stranded Dec 12–14, min-stay is the blocker, one click to relax inside gaps”) instead of asking anyone to read a report. Same flag powers CS save plays and the pacing digest.
How it works
Three small builds convert, two medium builds compound, and the campaigns start before any of them ship.
Conversion layer (weeks 2–4)
- Mirror button. Where a Vrbo Mobile or OneKey discount exists with no Airbnb twin (and the reverse), offer “add the Airbnb version” at discount save and on the discounts grid. The twin copies property scoping and Stay Date blackouts; the amount is Airbnb-fixed at 15% for Highly Rated Guest, so there is nothing to mis-enter. Announce once, with blackouts on all four promotions as the headline.
- Enroll-all with pre-filled blackouts. On the Airbnb channel page: “Enroll N listings. We protected each listing’s priciest nights this year and next, review or edit.” The suggestion resolves each listing’s forward nightly calendar the way the rate engine already does and takes the top N per calendar year at the host’s protection level, pre-checked in the Stay Date calendar the host already knows. Show the bounty-share credit before the click. Default the trial framing: on through Dec 31, host chooses in January. Two refinements from the calendar: with about 77 nights left in 2026, protecting 30 of them (39%) over-asks, so the host picks an annual protection level on a slider (protect my N priciest nights per year, 0–30, default mid-scale) and the current year prorates (at 15/year: about 4–5 nights for the rest of 2026, 15 for 2027, each year inside Airbnb’s 30-night cap). And per-property date sets matter: one shared global top-30 covers a median 37% of a listing’s own peak nights, a Florida-wide set the same, while the synchronized Smokies reach 73%. Default to metro-cluster shared discounts where coverage holds, show each property its coverage percent in the wizard, and give properties under a coverage threshold (say 60%) their own dated copy. Bounds discount count for PMs without giving up the nights that matter.
- One-click reactivate for deactivated discounts, same pre-fill. Plus default-on (opt-out) for newly connected or imported Airbnb listings while the pilot pays.
- One-button Airbnb launch. For Vrbo-only accounts (106 of the cross-sell accounts, holding 1,180 Vrbo listings, plus 78 in the reverse segment), the listings, photos, rates, and rules are already in OwnerRez. The work is a guided connect: pick listings, map the few Airbnb-only fields, publish, arrive in the wizard with Recommended pre-selected and enrollment on. Vrbo’s host-fee rise to 12% makes this quarter the cheapest diversification pitch we will get for a while, and every launched listing is next year’s Productivity base.
Nights machinery (weeks 1–8)
- Min-gap campaign (week 1, no build). The availability calculator already relaxes minimum stay inside gaps when the rule is set; 77% of cohort listings have it unset. Target the 5,808 listings with 1–3 night bounded gaps, starting with the 2,139 whose warehouse-flagged orphan nights cannot book at all while the rule is off: “You have N stranded nights in Q4. One setting makes them bookable.”
- Gap-night pricing (the one real pricing build). Recommended shape: a new criteria (criteria are the condition rows hosts already stack on discounts: listing site is, property in, stay dates between) on Rate Adjustment discounts, not a new property setting.
- Why adjusters: hosts already think in adjusters, adjusters fold into each night’s price in the rate calc (your memory is right: channel pushes already roll them into nightlies, while discounts proper go out as Airbnb pricing rules), and a criteria composes with site, property, and season scoping. Bands stack: “-20% when the arrival-or-departure gap is 1–2 nights, -10% at 3–4.”
- Blocker 1, the field gate: surcharge criteria currently reject the gap fields (Arrival Gap, Departure Gap, Arrival or Departure Gap) for every discount type. Allowing them for Rate Adjustment (and standard discounts) is a small deliberate change of the same kind Stay Date just went through.
- Blocker 2, nightly evaluation: quotes already evaluate gap fields against the proposed stay (unbooked nights adjacent to it), so direct works on day one. Channel pushes turn criteria into date ranges first, and gap membership is not a date property. The nightly build must compute each night’s open-run length from bookings; the availability calculator already walks gaps for min-stay rules and the warehouse stamps orphans, so wire one in, apply matching adjusters after spot rates (labelled in the rate tester), and bump the rates-change stamp on adjacent booking create/cancel where a gap adjuster exists.
- Semantics to decide in the card: stay-based (quotes) vs night-based (pushes) differ at the edges; document the push approximation instead of inventing a second criteria family.
- Fallback: a single “reduce by X% inside gaps up to Y nights” percent beside the min-gap rule on the property. Fewer moving parts; no per-channel amounts, no bands.
- Prerequisite pool split: warehouse orphans (7,811 nights) cannot book at any price until min-stay relaxes; gap pricing targets short gaps that can book and do not. Ship the pair, measure separately.
- Monthly ladder template. One click adds 7/14/28 tiers at sensible defaults, season-scoped so a cheap month cannot eat peak weeks. Ship as a template in the discounts flow plus the Florida campaign; no new pricing engine work, length-of-stay discounts already map to Airbnb’s long-stay rule.
- Recommended-discounts wizard. Most hosts will never compose a discount stack from a blank page, so stop asking them to. Wherever discounts are empty (new property setup, channel connect, the discounts page empty state), offer four doors: Import from Airbnb (pull the listing’s current Airbnb rule sets into OwnerRez discounts, the way listing import already builds promotions; scope the read path in the card), Recommended (one click installs the pack below for their market), Set up myself (today’s editor), None. Recommended defaults mined from what top-quartile Q4 2025 bookers in the cohort actually run, per market (correlation, labelled as such):
Market-specific exclusions belong in the wizard, not in host folklore: monthly tiers stay off where short-stay demand dominates (Smokies pattern) and where HOA or city rules already force 28+ night minimums the logic inverts (protect rate, skip the discount); Saturday-to-Saturday beach markets get the weekly tier aligned to the rental week and no Last Minute undercutting of it; event and ski markets pair the pack with suggested peak blackouts so discounts never touch the ten nights that carry the year. Provenance, so we can say it in product without blushing: every number above is OwnerRez data (config presence among top-quartile Q4 2025 bookers in the 14,688-listing cohort, tier amounts are medians of 3,300+ active tiers hosts run today). No third-party dataset needed for v1; AirDNA or PriceLabs market aggregates are an enrichment later, not a dependency. How the wizard knows a property’s market traits without asking: arrival day-of-week concentration and length-of-stay histogram from its own bookings (Saturday-to-Saturday shows up as Sat-arrival share plus exact-7-night share; no cohort metro exceeds 20% Saturday arrivals, so detect per property, never assume by label), lead-time profile, rate seasonality shape, channel mix, dynamic-pricing presence, and its own minimum-stay setting for forced-long-stay. Jurisdiction and HOA rules are the one thing we must not infer from an address; the wizard asks a single question there and otherwise defaults from observed demand. Industry-leading here is not more knobs: a default that is 90% right per property, a preview priced on the host’s own forward calendar (“at these settings your December gap nights price at $X, your protected nights stay $Y”), one-line evidence per default (“monthly on: top Florida bookers run it at 26% vs 16%”), seasonal differencing built in, and one-click undo with a 30-day report card. Airbnb’s own host nudges and PriceLabs recommendations set the bar on specificity; ours wins on being cross-channel and honest about direct. This wizard is the enrollment surface too: Highly Rated Guest with pre-filled blackouts is one more pre-checked row in Recommended for Airbnb listings.
Market What winners run (top quartile vs rest) Recommended default Florida beach / Orlando (5,146) Monthly 26% vs 16%, Early Bird 23% vs 16% 7/14/28 ladder on at -10/-15/-20% (the medians hosts run), Last Minute on, Early Bird on Gulf / SE coast ex-FL (3,054) Monthly 32% vs 22%, weekly+ 40% vs 29% 7/14/28 ladder on TN Smokies (1,364) Monthly 8% vs 10% (no edge), weekly+ 30% both Weekly tier + Last Minute + Early Bird; monthly off by default Mountain (664) Monthly 28% vs 16%, Early Bird 20% vs 7% Early Bird forward, monthly on for shoulder seasons Desert / West (905) Early Bird 21% vs 12%, monthly flat Early Bird + Last Minute, weekly tier on - Block prompts. Weekly surface, per listing: blocked nights in the next 60 days with one-click release of selected ranges. Cheap, and it respects owner intent by asking instead of assuming.
- Pricing cockpit (generic). One war-room surface per account, built on the calendar warehouse plus a nightly
Bookingrollup, with four readouts hosts can act on without a report: (1) pacing, nights and GBV booked for future windows vs the same point last year, split by channel with direct alongside Airbnb, Vrbo, and Booking.com; (2) unbookable inventory in dollars, warehouse orphan nights and looming stood blocks priced at each night’s resolved Rent, classed as imported-rolling, seasonal-not-yet-opened, or suspicious, each with its one-click fix (relax min-gap, release or confirm the block, extend the booking window); (3) discount gaps against the wizard’s market pack; (4) rate-compare deltas once the compare tool ships. Every card carries expected nights and dollars so the page reads like a table of priced bets, not homework: cost to pull the lever on the left, expected payout on the right. Surfaces: the cockpit page, a Monday digest with only what changed, the calendar ribbon (the multi-property calendar strip; its day model already carries GapType Orphan, so orphan nights can render loud there with a count badge per property), and our internal pace view (enrollment vs the 10% gate; cohort nights and GBV vs the Q4 2025 curve). Airbnb’s Impact Data expectation is met by the same page filtered to Airbnb, not by an Airbnb-branded scoreboard. A bizdev read: loss in dollars outperforms percentages, fixes sorted by expected value get clicked, and a weekly delta email turns a dashboard nobody visits into a habit. - Multi-channel rate compare (scoped in Q4, built to last). Extend the rate tester: same dates and guests, guest-total and host-net side by side for Airbnb, Booking.com, Vrbo, and direct, with adjuster, fee, and tax lines separated. Warn when the Airbnb guest total sits more than ~5% above the cheapest major channel for the same stay. This retires Airbnb’s price-competitiveness ask with a tool hosts keep using after the incentives end.
Sequence
| When | What ships |
|---|---|
| Week 1 (Oct 15–22) | Five target lists cut; min-gap + block + Florida monthly campaigns start; manager packet template; asks sent to Dianna (denominator, tier table, counting rule, co-marketing). |
| Weeks 2–4 | Mirror button + announcement; reactivate; enroll-all with pre-filled blackouts; cockpit v1 (pacing + orphans); self-serve sprint to the 10% gate. |
| Weeks 4–8 | Manager payloads enroll at $20; gap-night pricing ships behind the same campaigns; caption sprint and parity rescue as door openers; compare tool scoped and started. |
| December | Retention mode: disenrollment saves, beat-your-Q4-2025 pushes, protect the gate to Dec 31. No new launches after mid-December. |
What could go wrong
- Hosts pad prices 15% to neutralize the discount. The predictable gaming move, and exactly what Airbnb’s price-competitiveness push exists to stop. We never teach it; we watch adjuster changes on newly enrolled listings and keep the pitch on guest quality and conversion. If padding shows up, pacing and rate data make it visible per listing.
- Managers say yes, owners say no. A manager often cannot discount owner revenue alone. Mitigation is the owner one-pager and land-and-expand pilots on manager-owned or pre-approved listings first.
- Disenrollment before Dec 31 drops us under 10% and downgrades every $20 listing. Mitigation: cross early with margin, alert on every switch-off, same-day save attempt, and report the buffer (enrollments above the gate) weekly.
- The nights incentive may count stay dates, not booking dates. Then January snowbird bookings taken in November earn nothing for Q4. The monthly push still makes hosts money, but we should size the Q4 claim honestly until Dianna confirms the rule.
- Gap pricing churn. Every adjacent booking change re-prices a window. Bound it: only gaps up to a set length, only inside the normal push cadence, and a kill switch per account. Dynamic-pricing users must see the layering in the rate tester or trust erodes fast.
- Contract limits on bounty share. We have shared before, but confirm this pilot permits host pass-through before printing a dollar figure in any UI.
- Quoting biased stats. Enrolled listings book more partly because better operators enroll. Use per-host backtests (“your own last year”) instead of population claims; managers will check.
Other options considered
- Email and banner only. Cheapest, and probably enrolls a few hundred listings. It leaves the $20 gate to luck and the stranded nights stranded. Rejected as the whole plan; kept as the announcement layer.
- Mass default-on enrollment without a host click. Fastest route past 10% on paper. It discounts host revenue without explicit consent, invites chargebacks of trust we cannot re-earn, and likely violates the spirit if not the letter of the pilot. Rejected. Default-on survives only at connect/import, clearly shown, one click to remove.
- Build everything, including a full autopilot pricing agent. Tempting and wrong for an 11-week window: anything shipping after mid-November has six weeks to pay back. The compare tool is the one deliberate exception because its value outlives the incentives.
Who does what: three tracks
Product builds the doors, Paul Hall fills the top of the funnel, Steve walks named accounts through them. Host-facing copy in every track is channel-agnostic: book more overall; Airbnb wins its share.
| Track | Owner | Deliverables |
|---|---|---|
| 1. Build | Chris (product) + Engineering; Shawn Harwell sequences releases | Mirror button; enroll-all with pre-filled blackouts; one-click reactivate; default-on at connect/import; gap criteria on rate adjusters plus the nightly price build; recommended-discounts wizard (import / recommended / myself / none); pricing cockpit; one-button channel launch assist; multi-channel rate compare (scoped now, built to last); segment refresh pulls. Bar: conversion layer live by early November, December quiet. |
| 2. Mailings + partnership | Paul Hall (Marketing & Partnerships) | Announcement bundle (blackouts on all four promotions, mirror, wizard); geo-kit mailings (Florida monthly, Smokies holiday short-stay); pacing digest emails to the 3,435 cohort accounts; Vrbo-fee-moment expansion campaign; newsletter and in-app banners for the 1,232 / 163 / 166 segments; Airbnb asks and co-marketing (denominator, tiers, counting rule, host-facing placement, Impact Data expectations); parity-rescue comms for the 144. Paul Waldschmidt approves bounty-share spend. |
| 3. Outreach | Steve (CS Lead) + CS/sales | Top-200 manager packets and land-and-expand pilots (five listings, 30-day report, portfolio ask) with the owner one-pager; resurrection calls to 166 accounts; cousin and mirror follow-ups where self-serve stalls; min-gap turn-on assists for the 2,139 listings with orphans and the rule off; block-release follow-ups and the Christmas-week blank-block audit; disenrollment saves through Dec 31. |
Hosts and managers see one-click offers with their own numbers, a pacing view across all their channels, and fix lists for their listings. Nothing changes price, enrollment, or channel presence without their click.
Decisions needed
- Bounty share: how much, what form? Recommendation: a per-listing OwnerRez credit, shown in the enrollment flow, sized after confirming the pilot permits pass-through. The $15/$20 ceiling makes $5–$10 per listing easy to defend.
- Gap-night pricing: build in Q4? Recommendation: yes, it is the largest measurable nights pool and the mechanism is bounded. Needs an owner and a mid-November ship bar; later than that, run the campaign without it.
- Default-on at connect/import? Recommendation: yes, with suggested blackouts attached and removal one click away, only while the pilot pays.
- Who owns the blitz list and the weekly gate report? Recommendation: CS runs accounts, product runs the ticker; one shared view of enrollments above the gate.
- What do we ask Airbnb for, in what order? Recommendation: eligible denominator + today’s enrollment rate; Productivity tier table; booked-vs-stay counting rule; bounty pass-through confirmation; host-facing co-marketing and quotable search-impact data.
- Gap pricing mechanic: criteria on rate adjusters, or a property percent? Recommendation: criteria on rate adjusters (composes, per-channel amounts, bands), accepting the nightly price build and the documented stay-vs-night semantics split. Property percent only if that build cannot make mid-November.
- Wizard defaults: mined from winners, or hand-set? Recommendation: ship the market table in this page as v1 defaults (labelled correlation), then re-mine quarterly from top-quartile configs so Recommended stays evidence, not opinion.
- Blackout suggestion defaults: slider level and sharing granularity? Recommendation: annual slider (default 15 nights/yr, current year prorated), metro-cluster shared discounts with a 60% own-peak coverage threshold below which a property gets its own copy.
- Cockpit scope for Q4: pacing only, or pacing plus unbookable dollars? Recommendation: both, because the warehouse already computes orphans and Rent nightly; discount gaps join when the wizard ships, compare deltas when the compare tool ships.
Data and method
Airbnb’s eligible-listing sheet (14,700 listing IDs, Oct 6) was matched to production on the read replica via LinkedAccountProperty.LinkedAccountToken: 14,693 matched (7 unmatched), resolving to 14,688 properties across 3,435 accounts. Cohort figures join Property, PropertyListing, Address, Booking (arrivals Oct 2025–Jan 2027), Surcharge with criteria scoping (site and property conditions honored), SurchargeAmount thresholds for length-of-stay tiers, SpotRate/SeasonalRate calendars, and File photo aggregates. Q4 windows clip booking nights to the window; gaps are open runs bounded by a booking or block on both sides. Platform promotion counts are active, non-deleted discounts joined to their listing-site criteria (Vrbo = site 2, Airbnb = 15, Booking.com = 20; discount types from SurchargeType). Analysis scripts and extracts live with the author on request; every figure in this page traces to those extracts. Pacing uses surviving bookings only (bookings later canceled are invisible to the as-of reconstruction, identically in both years). Block release timing uses deleted/canceled stamps on block rows; reason buckets are keyword groupings of free text and undercount every named purpose. Dynamic-pricing dominance uses SpotRate.LastApiApplicationId over the next 365 days. Channel presence joins live LinkedAccountProperty rows (Airbnb by host id, Vrbo by listing URL, so Vrbo counts are floors) and last-12-month bookings by listing site. Warehouse figures read CalendarWarehouse nightly rows (resolved rent, effective rules, orphan flags) for Oct 16–Dec 31, 2026, covering 97.1% of the cohort. Segment CSVs (mirror-ready, expansion, cousins, deactivated, min-gap-off, discount gaps, top-200 packets) were cut from the same extracts. GBV uses Booking.TotalAmount pro-rated by nights inside the window, USD only (null-currency bookings excluded, so dollars understate; ratios are the usable output). True cancellation and Instant Book values resolve the PropertyListing override first, then LinkedAccount RelatedItems defaults, joined to listings through the Airbnb host id (94.5% matched; the replica user cannot select LinkedAccount.ID). Dynamic-pricing dominance is the dominant SpotRate.LastApiApplicationId per listing over the next 365 days, and peak-overlap figures come from 295 sampled full spot calendars (Tennessee n=26, read as directional).
Caveats that change how to read three numbers. One, several Airbnb fields (Instant Book category, cancellation policy, prep days, lead time) are unset for most listings in our tables and take channel defaults, so explicit-setting rates are floors. Two, today’s platform-wide Highly Rated Guest listing coverage is estimated from the Oct 14 launch-labor dry run (664 discounts covering 4,190 listings across two promotion types); the exact enrolled count and the contract’s eligible denominator must come from Airbnb. Three, the Productivity payout tiers and its night-counting rule (booking date vs stay date) were not in the material provided; night yields above are therefore stated in nights, not dollars.
Private draft for internal discussion. Figures pulled Oct 15, 2026; Q4 windows move daily. Nothing in this page has been shared with Airbnb.