Airbnb's New House Rules and Vrbo's 12% Commission: The Direct Booking Response

Two changes landed inside a month of each other, and together they close the loop that most vacation rental operators have quietly depended on for years: book the guest on an OTA, capture their contact details during the stay, and market them into a direct booking next time. Vrbo raised the cost of the first step. Airbnb restricted the second.

Key Takeaways: Vrbo moved to a flat 12% commission on 29 October 2026, which for software-connected managers is a jump from roughly 5%. Airbnb’s new house rules take effect 20 November 2026 and prohibit soliciting guest contact details, ID documents and financial data, restrict rental agreements on short stays, and bar requiring a third-party app or website registration to gain entry or access a critical amenity. Airbnb has built the policy into the Listing editor, which scans listing text as you save it and escalates from a red flag to reduced search visibility and ultimately account removal. Neither change bans guidebooks, in-unit branding, QR codes or direct marketing. The line is offer versus gate: if the guest never engages and nothing breaks, you are fine. The upside is that the replacement converts better than what it replaces. A guest who opens your guidebook for the door code, claims an arrival-day treat and screenshots your local shortlist is worth more than an email address extracted at check-in, because they wanted something from you first.

---

What Actually Changes on Airbnb on 20 November

Airbnb is consolidating house rules into one place in the listing editor and removing the separate checkout instructions tab. That part is cosmetic. The substantive part is what you may ask of a guest.

What the policy now says. Each of these is something you could do last month and cannot do on a new booking next month:

• Guest contact details are off limits. Airbnb's language covers "contact information, such as their phone number, home address, or email address", at any point in the booking process or the stay.
• ID documents are off limits to collect or keep. You may still check a guest's ID in person at check-in if you disclose that in the requests section, but you may not retain a copy, digital or physical, outside the narrow exceptions.
• Rental agreements are restricted. Contracts survive only for a legal requirement, a documented HOA or community association rule, monthly stays, Airbnb Luxe listings, and coverage of optional amenities. A standard short stay is not on that list.
• Requiring registration to get in is prohibited. Hosts may not "require use of a third-party app or registration on another website to gain entry to a listing or access a critical amenity." Note what this restricts: what you can require. It says nothing about what you can offer.
• Checkout chores are out. Stripping beds, running the dishwasher, hauling trash to the curb.
• Fines and surprise fees are out. Any penalty framed as a charge for breaking a rule.
• The listing is the entire rulebook. You may not add a rule or a request after booking, and Airbnb is explicit that this covers message, sign, and in person alike. A sign may remind a guest of a rule already in the listing. It may not create one.
• Exceptions must be disclosed up front. Where one genuinely applies, hosts "must clearly disclose what information is required of guests, and why, in the requests section of their listing page so that guests can review it before booking."
• It binds new bookings only. The rules apply to reservations booked on or after 20 November 2026. Earlier bookings stay under the old policy, so you will run two rulebooks in parallel for months.

What To Clear Before 20 November

This is the housekeeping half, and it is the smaller half. Work down it once, then spend your real effort on the list after it. Each line here is something to find and remove, or find and move.

• Strip email requests from everything. Listing text, the requests section, saved message templates, and automated pre-arrival sequences. The templates are the ones people forget.
• Strip phone number requests. Including the common “text us your number and we will send the gate code”.
• Remove ID photo and document requests. If you check ID in person, disclose it in the requests section and keep no copy.
• Pull rental agreements off short stays. Keep them only where one of the listed exceptions genuinely applies, and be honest about whether yours does.
• Get the door code out from behind any download. If entry depends on an app install or a site registration, that is the single clearest violation in the whole policy.
• Audit your WiFi splash page. If a guest cannot reach the internet without submitting an email or phone number, fix it before the date, not after.
• Delete checkout chores. Remove them from the listing instructions and from any in-unit signage that repeats them.
• Remove fines and penalty framing. That includes charges presented as deterrents rather than as fees.
• Re-read your in-unit signs, welcome books and QR codes. A sign cannot create a rule or ask for data. Most operators forget these exist, because nobody edits a laminated card.
• Move every surviving rule into the listing itself. After 20 November that is the only place a rule can live.
• Disclose any genuine exception in the requests section. State the reason too, so guests see both before they book.
• Check which reservations are actually affected. Bookings on or after 20 November only, so do not rewrite the guest experience for stays already on the books.

What To Build Instead

Here is the part worth being cheerful about. Everything above is deletion. None of it grows anything, and if you stop there you have a compliant listing and the same conversion problem you had in October.

The opportunity is that most operators were never really converting OTA guests anyway. They were collecting addresses. An email harvested at check-in from someone who just wanted the door code is a row in a database, not a relationship, and it opens at a rate that reflects exactly that. What replaces it is better, because it starts with the guest wanting something.

• Move the access details into the guidebook, ungated. Door code, gate code, pool code, parking spot, bin day, the trick with the second lock. No signup in front of any of it. This is the thing that makes guests open the guidebook in the first place, and you cannot run any of the rest without that first open.
• Build the arrival-day offer. One genuinely nice thing, claimable from inside the guidebook, for a cost you can absorb. Bakery goods delivered the morning of arrival. A bottle of local wine. Coffee from the good roaster two streets over. It costs you a few dollars a stay and it is the single most reliable reason a guest will give you an email address, because they are getting something they actually want on a day they are already excited.
• Build the experiences shortlist. Not twenty links. The six things you would tell a friend. Which beach access has parking, which restaurant needs booking a week out, where the farmers market actually is on a Saturday. Guests screenshot this, which means they open it repeatedly, which means the offers beside it get seen more than once.
• Write the directions people actually need. How to reach the beach on foot, where the trailhead parking fills by nine, which door the pool key opens and what the gate hours are. This is the content guests return to mid-stay, and mid-stay is when they are deciding whether they would come back.
• Put a QR code on the welcome card, not next to the lock. Beside the kettle, on the welcome card, inside the cupboard door. It should open the guide, the offer and the local list. Nothing is withheld from a guest who ignores it.
• Get contact details from the WiFi, not from the booking. Let guests reach the internet without filling anything in, then put the opt-in right beside the connect button: the local guide, the arrival-day offer, next year’s rate. It reaches every adult in the unit rather than only the person who booked, and SendSquared matches each opt-in to the active PMS reservation for that property, so an occupant who was never on the reservation still lands on the right stay instead of in a spreadsheet.
• Publish the returning-guest rate where a happy guest will see it. Day four of a good stay, inside the guide they are already reading. This is the warmest audience you will ever address and almost nobody is addressing it.
• Put the upsell menu somewhere guests can act on it. Late checkout, early check-in, mid-stay clean, crib, boat, golf. Each one is revenue, and each one is a transaction that legitimately gives you a name, an email and a card.
• Ask for the review and the survey after checkout. Send both through Airbnb messaging, with nothing gated behind either of them.
• Decide what a first-time OTA guest is worth to you. You cannot tell whether 12% was a fair price for them without that number, and without it the rest of this list is decoration.

What links every line is the same inversion. You are no longer asking a guest for data because you need it. You are giving them something and letting them decide to stay in touch. The list that comes out the other end is smaller and worth considerably more, because every person on it raised their hand.

The Scanner Airbnb Is Deploying

This is the part most coverage skips, and it changes how seriously to take the checklist above.

Airbnb has built the policy into the Listing editor itself. When you save a listing request, the editor scans the text and underlines problem phrasing in red with a short explanation of what is wrong. It is not a passive policy document you might get away with ignoring. It reads what you type.

The editor currently recognises seven warning categories: chores, fines, mandatory fees, local fines, contracts, personal info, and off-platform registration.

When operators tested it, the example that got flagged under off-platform registration was exactly the sentence thousands of listings contain today:

“Please download our guest app to get your door code.”

That is the whole problem in one line. Not the app, not the guidebook, not the branding. The door code sitting behind the download.

Two things worth knowing about how the scanner behaves. First, flagged text can still be saved. The flag is not a hard block, and Airbnb has said flagged requests then go on for further review, so saving past a red underline is not approval, it is a queue. Second, it produces false positives: testers found it flagging a bear-proof bin request that Airbnb’s own policy offers as an example of a reasonable ask. Whether a human reviews flagged items or the decision stays automated is a question the industry has put to Airbnb and Airbnb has not yet answered publicly.

The practical reading: treat a red underline as a signal to rewrite, treat a saved flag as unresolved rather than cleared, and do not assume a false positive will resolve itself in your favour.

How Airbnb Monitors This, and What Happens If They Find Out

Monitoring runs on three surfaces, and operators tend to only think about the first.

The listing text, scanned automatically by the editor as described above. This is the surface you control and the easiest to clean.

The guest. Airbnb has been showing guests what a listing asks of them, including guest-facing alerts on listings that carry flagged requests. A guest who is asked for something the policy no longer permits now has both the context to recognise it and an obvious route to report it. For anything happening inside your unit, a sign, a QR code, a welcome book, the guest is the detection mechanism. There is no crawler in your living room. There is a person with a phone who has just been told what you are allowed to ask.

The message thread. Everything you send through Airbnb messaging is on Airbnb’s infrastructure. Asking for an email in a message is not more private than asking in the listing, it is simply less visible to you.

The consequences escalate, and they are worth knowing in order:

1. A red flag in the editor when you save. Cosmetic, free to fix.
2. Further review once a flagged request is saved anyway.
3. A guest-facing alert on the listing, the public one, telling prospective guests your listing includes a flagged request. This costs you conversion before it costs you anything else.
4. Reduced search visibility for listings that keep violating. This is the expensive one, because it is invisible from your side. Bookings simply stop arriving and nothing tells you why.
5. Listing suspension.
6. Account removal for repeat offences.

Step four is the reason not to gamble. A delisting is at least legible. A quiet ranking penalty on a portfolio looks exactly like a soft market, and operators have spent entire seasons blaming demand for an enforcement action.

Weigh that against the upside, which is one email address you could have obtained a legitimate way. It is a bad trade on expected value, and it is a worse trade on information: you will not find out you lost until the quarter is over.

What Changes on Vrbo

Vrbo notified hosts on 29 September 2026 that it was moving to a flat 12% commission, effective 29 October.

The impact depends entirely on how you were paying before:

Host type	Before	After
Software-connected (PMS or channel manager)	about 5%	12%
Pay-per-booking	about 8% (5% + 3% processing)	12%
Legacy annual subscription	annual fee + 3% processing	12%

For professional operators the first row is the one that stings. Going from 5% to 12% is not a 7 point increase, it is a 140% increase in what you pay Vrbo per booking. On a portfolio doing one million dollars of Vrbo revenue, that is roughly fifty thousand dollars of commission becoming roughly one hundred and twenty thousand.

Two qualifications. Vrbo states the rate ranges from 12% to 15% depending on region and brand, with VAT or GST on top where applicable, so check your own dashboard rather than assuming the headline number. And taxes and refundable damage deposits sit outside the commission, though they still carry a 3% payment processing fee.

Industry estimates put software-connected listings at roughly 55% of Vrbo’s global supply and about 51% in the US, which is why this landed as hard as it did. The majority of Vrbo’s professional inventory is in the group whose rate more than doubled.

The Thing Most Operators Will Get Wrong

The instinct is to look for the workaround. Find the phrasing that still lets you ask for the email. Move the request from the listing to the door sign. Put the agreement in the welcome message instead of the booking flow.

Do not do this. Airbnb has shipped a rules checker that scans listings against the new policy, and the enforcement surface is the listing itself. The downside of being wrong is suppression or delisting, against an upside of one more email address. That is a bad trade on expected value alone, before you get to whether you want your guest relationship to begin with a guest feeling processed.

There is also a simpler reason: the workaround solves the wrong problem. If your direct booking strategy depends on harvesting contact details from guests the OTA sent you, your strategy was always running on someone else’s permission. The policy change did not break it. It revealed that it was borrowed.

What Still Works, Precisely

The useful question is not what Airbnb forbids. It is which guest touchpoints you actually own. Those are unaffected by an OTA policy because they do not happen inside the OTA relationship.

Guidebooks, delivered rather than gated. This is the distinction that matters and it is worth being exact. Airbnb is not banning digital guidebooks. It is banning the pattern where a guest must register somewhere else to get in the door. A guidebook that you send through Airbnb messaging, that opens without a signup, and that carries your brand, local recommendations and upsells, is still entirely available to you. What has to go is the registration wall in front of the check-in instructions.

On-property WiFi, where the opt-in is real. Your network is yours, and a guest connecting to it is not a booking condition, which makes a captive portal the most durable first-party capture point most operators have. It reaches the whole party rather than only the person who booked, which is why WiFi capture outperforms every other on-property channel.

A guest should not have to hand over an email address to get online. That is worth saying flatly, because the standard captive portal does exactly that: it puts a form in front of the internet and calls whatever comes out of it a lead. Nobody should be paying for WiFi with their contact details in a unit they have already paid for. From 20 November that design also has a policy problem, because WiFi is plausibly a critical amenity and Airbnb now bars requiring registration to reach one.

The better pattern is not complicated. Let guests connect. Put the offer beside the connect button instead of in front of it: the local guide, the late checkout, the returning-guest rate, whatever is genuinely worth an email address. Fewer people will opt in. The ones who do actually want to hear from you, which is the only kind of list that is still worth sending to a year later.

Inbound calls. A guest who phones you is already your contact, with no OTA intermediating. Most operators cannot say how many calls they take, how many go unanswered, or what those calls were worth. Call tracking and AI voice turn the phone from an unmeasured cost into an attributable acquisition channel, and after-hours coverage converts the calls you are currently losing to voicemail.

Your own booking path. Everything above is capture. This is conversion. A direct booking engine is what makes a captured guest worth capturing, and at 12% commission the arithmetic on every direct booking just improved by seven points.

The guest record underneath it all. Capture is only useful if it accumulates. A guest CDP unifies reservations from your PMS with WiFi opt-ins, call history and messaging into one profile per guest, lifetime value makes the repeat-guest economics visible, and segments turn that into an audience you can actually target. See vacation rental CRM for how the layers fit.

In-Unit Branding and QR Codes: Where the Line Actually Sits

The unit is the one space in the whole guest journey that is unambiguously yours. Airbnb mediates the booking, the messaging and the payment. It does not own your coffee table.

That matters more now, not less. If you cannot ask for contact details in the booking flow, the physical stay becomes the place where a guest decides whether you are a brand they would book again or an anonymous key handover. Most operators have spent years optimising listing photos and almost no time on what the guest actually sees once they are standing in the room.

What in-unit materials may still do:

• Carry your name, logo, colours and voice on the welcome card, the coffee packaging, the bathroom amenities, the door mat, the printed guide
• Remind guests of a rule that is already published in the listing
• Recommend restaurants, beaches, trails, the good coffee two streets over
• Offer something: late checkout, a mid-stay clean, a local experience, next year’s returning-guest rate
• Point to a guidebook, a concierge chat, or a loyalty programme that a guest may ignore with no consequence

What they may not do:

• Create a rule that is not in the listing
• Ask for an email, a phone number, an ID or a signature
• Send a guest somewhere to register in order to get the WiFi password or the door code
• Charge, fine, or threaten a charge

QR codes sit exactly on this line, and the line is entry versus offer.

A QR code taped beside the door reading “scan to register for your access code” is the clearest violation available to you. It is off-platform registration gating entry, and it is the same sentence the listing editor already flags, just printed on vinyl.

A QR code on the welcome card reading “scan for our local guide, the beach parking trick, and the number for the good taxi” is a different object entirely. Nothing is withheld from a guest who ignores it. It creates no rule. It asks for nothing. It is an offer, and offers are untouched by this policy.

The design test is a single question: if the guest never scans it, does anything break? If the answer is yes, you have built a gate and you need to remove it before 20 November. If the answer is no, you have built an offer and you can put it in every unit you operate.

A few that pass that test cleanly:

• The local guide code, scanning straight into your guidebook with no signup wall, branded as yours
• The upsell code, going to late checkout, early check-in, a crib, a mid-stay clean, a boat or golf booking
• The service code, scanning into a chat thread that reaches your team, which is more useful than a phone number on a fridge magnet and is also how you find out about the broken dishwasher before the review
• The returning-guest code, offering next year’s direct rate to someone standing in the unit who is already, demonstrably, your ideal customer

That last one is the quiet winner. A guest on day five of a good stay is the warmest audience you will ever have, and until 20 November almost nobody was asking them anything, because everybody was busy harvesting emails at check-in instead.

The Guidebook Is the Conversion Engine

Worth separating the compliance point from the commercial one, because they get conflated. The compliance point is that a guidebook cannot gate entry. The commercial point is that a guidebook is the best conversion surface you have, and the policy change barely touches it.

Think about what a guest actually does. They arrive late, tired, with a car full of bags, and they need the door code and the WiFi password. If both live in your guide, they open it. That first open is everything, because it is the only moment you are guaranteed their attention, and it costs them nothing.

So put the useful things in and let the guide earn the second open:

The access details. Door code, gate code, pool code and hours, parking spot, bin day. Ungated, because that is both the rule and the reason they came.

The arrival-day offer. This is the mechanic worth building deliberately. One good thing, claimable from inside the guide, on a budget you barely notice: bakery goods delivered the morning they arrive, local wine, a coffee voucher. A guest gives you an email because they want the croissants, which is a completely different transaction from being made to register before the lock opens. The first is marketing. The second is the thing Airbnb just banned. Nothing is withheld from a guest who ignores the offer, which is what keeps it on the right side of the line.

The local shortlist. Six recommendations you would give a friend, not a directory. Which beach access has parking. Which restaurant needs a week’s notice. Where the market is on Saturday morning. Guests screenshot this and come back to it, and everything beside it gets seen again each time.

The practical directions. Walking route to the beach, where trailhead parking fills by nine, which door the pool key opens. Mid-stay content, read at exactly the point a guest is forming an opinion about whether they would return.

The returning-guest rate. On day four of a good stay, inside something they are already reading. The warmest audience in hospitality, and almost nobody addresses it.

What We Automate

A few of these are tedious by hand across a portfolio and are the parts SendSquared guidebooks handle directly.

Guidebooks are reservation-based, not generic. Each one is tied to units through your PMS, so a guest’s link opens with their own check-in and check-out dates, guest count, unit details and photos, with merge fields putting their name in the copy. You write a property once rather than maintaining a PDF per booking.

Time-based access windows. You set when a guidebook opens and closes relative to the reservation, for example seven days before arrival and three days after departure, with a separate window controlling when the property address appears, and per-block timing underneath that. Expired guidebooks show clean messaging rather than stale content. In practice this is how the arrival-day offer only shows up on arrival day without anybody scheduling anything.

Conditional visibility. Blocks can show or hide per guidebook or per block with AND/OR logic evaluated at render time, so the pool section only appears for units with a pool and the ski content only appears in winter, from one template.

Share the Trip. One guidebook the whole party can open, which matters for the same reason WiFi capture matters: the booker is one person and the group is four, and the other three are also people who might book next year.

It is tracked. Engagement and upsell conversion are measured end to end, so you can see which blocks get opened, which offers convert, and which properties have a guide nobody reads. A printed welcome book cannot tell you any of that.

WiFi logins link themselves to the reservation. On the capture side, WiFi opt-ins are matched automatically to the active PMS reservation for that unit using check-in dates and property assignment, so an occupant who was never on the booking still resolves to the right stay, the right property and the right guest record rather than landing in a spreadsheet as an orphan.

On QR codes specifically, to be accurate about what is and is not a product feature: a code pointing at a guidebook link is just a printed URL, and you can generate one for free anywhere. What we do add is trackable QR placement on printed postcards and letters sent from the CRM, which is a different job: reaching past guests at home rather than current guests in the unit.

Where Guest Details Legitimately Come From Now

The question operators keep asking is how they are supposed to get an email, a phone number, or a real full name when the OTA will not give it and the policy forbids asking for it. The framing is the problem. You do not need Airbnb’s guests to hand over data. You need guests who are transacting with you rather than with Airbnb, and there are more of those moments than most operators use.

Run through the realistic scenarios.

The guest books direct. You are the merchant of record. You get the full name, email, phone and billing address, because you need them to take the booking. No OTA policy touches this, because no OTA is involved. Everything else in this list is a way of making more of these happen.

The guest calls you. Caller ID gives you a real, working phone number attached to a person who wants something. The conversation gives you the name. Call tracking turns that into an attributable record instead of a note on a pad, and AI voice catches the after-hours calls that currently go to voicemail and then to a competitor.

The guest buys an add-on. Early check-in, a crib, a mid-stay clean, a boat rental, a dinner reservation you are taking a cut of. The guest is purchasing from you, so you legitimately hold the transaction details, and the policy explicitly keeps contracts available for optional amenities. This is the single most under-used route in the whole list, and it has the pleasant property of making money on the way through.

The guest joins the WiFi and chooses to opt in. Covered above, with the one rule that matters: let them connect either way.

The guest scans a code because they wanted what was behind it. The guidebook, the guide to the area, the returning-guest rate. They gave you an email because they wanted the thing, which is also why that email will still open a message next spring.

The guest answers a post-stay survey. Sent through Airbnb messaging, answered voluntarily, no gate in front of the stay. People who had a good time will tell you so, and a reply is a contact.

The guest comes back. The second stay is where it compounds. Once someone has booked direct even once, you hold the full record, and lifetime value starts accruing against a guest rather than against a reservation.

On full names specifically, since this is the field operators complain about most: OTAs commonly pass a first name only, and no amount of asking changes that. Real full names come from transactions. A direct booking, a paid upsell, a signed optional-amenity agreement, a payment. If the name matters to you, the answer is to give the guest a reason to buy something from you directly, not to find a sharper way to ask.

What ties this together is that none of these records are much use scattered across a PMS, a phone system and a spreadsheet. A guest CDP is what makes a WiFi opt-in, an inbound call, an upsell purchase and a direct booking resolve to one person rather than four, and segments are what let you act on the result. See vacation rental CRM for how those layers sit together.

The Honest Economics

Keep both platforms. The OTAs still reach demand you cannot reach alone, and they still fill nights that would otherwise sit empty. Leaving is a worse decision than paying.

What changes is how you account for them. At 12%, Vrbo is no longer a distribution channel that happens to cost something. It is paid acquisition, and it deserves the same scrutiny you would apply to any other paid channel: what does a first-time guest cost, what fraction return, and what are they worth when they do. An operator who can answer those three questions can price OTA dependence deliberately. An operator who cannot is simply paying whatever the platform decides next.

The repeat guest is where this resolves. A first stay booked through Vrbo at 12% is acceptable if the second stay comes direct. It is expensive if every stay pays the toll. Guest loyalty and repeat booking programmes and gap night fill are how that second stay actually happens, and both run on first-party data you gathered legitimately on your own property.

Two platforms changed their terms in the space of a month, with no consultation and no appeal. That will happen again. The only durable response is to own enough of your demand that the next announcement is an inconvenience rather than a crisis.

If you want to see what your current channel mix costs and where direct could realistically take it, book a demo and we will go through it with your numbers.