Log in or sign up

Loading your account…

A laptop on a desk showing an abstract dashboard with a pie chart dominated by one large colored segment

For investors

Underwriting Platform Risk: One Channel, One Algorithm

How a rental investor can price the risk of one booking platform into a deal: revenue concentration, ranking and fee changes, suspension, and haircuts.

Category

For investors

Written by

The JekkyStay team

Published

· 11 min read

Quick answer

Platform risk is the chance that one booking site changes its fees, its search ranking or your account status and cuts your revenue. To underwrite it, measure how much revenue depends on that one channel, then model a fee increase, a ranking drop and a lost-booking period as separate scenarios before you buy.

This guide is for investors analyzing a deal. Our owner guide on the real risks of relying on Airbnb covers the day-to-day side; here the question is how those risks belong in a pro forma.

This is general information, not tax, legal or investment advice.

Key takeaways

  • Platform risk is the chance that one booking site's rules, ranking or fees cut your revenue, and it grows with the share of bookings that site brings.
  • Airbnb's Terms say it can change service fees at any time with notice, and its fee page says most hosts are being moved to a single host-only fee of about 15.5%.
  • If a listing is removed, Airbnb says reservations with check-in after the removal date are canceled, so a model should include a lost-booking scenario, not only a slow month.
  • Underwrite platform risk with explicit haircuts: a fee-change case, a ranking-drop case and a short outage case, and check that the deal still covers its debt in each.
  • Diversify channels and build a direct guest list, but count their cost too; a second channel and a website are not free.
In this guide
  1. What platform risk is, in underwriting terms
  2. Measure the concentration first
  3. Fee changes: the easiest risk to model
  4. Ranking changes: the hardest risk to see coming
  5. Suspension and removal: the tail risk
  6. Refunds and rules that move money after the booking
  7. How to haircut a projection for platform risk
  8. Reducing the risk, and what that costs
  9. Questions to ask before you buy
  10. The bottom line

What platform risk is, in underwriting terms

Platform risk is a concentration risk: the more of a property's revenue comes through one channel, the more one company's decisions move your income.

Most pro formas treat revenue as a function of the market: comparable homes, seasonal rates and booked nights. That is right as far as it goes. But for most short-term rentals, the path from a guest's search to your bank account runs through one or two platforms, and each one sets four things that change your numbers:

  1. The fee it takes from each booking.
  2. The ranking that decides whether a guest sees your listing.
  3. The rules on refunds, cancellations and guest complaints, which decide what you keep.
  4. Your account status, which decides whether you can take bookings at all.

A lender or partner looking at a deal will ask how sensitive debt coverage is to occupancy. A careful one will also ask how sensitive it is to the platform. If 90% of bookings come from one site, a single policy change is a market event for that property, even if demand in the city has not moved at all.

Measure the concentration first

Before you can price the risk, you need to know how big it is, and that starts with where the bookings come from.

For an existing rental you are buying, ask the seller for twelve months of booking data by channel: nights, gross revenue and payouts for each platform and for direct bookings. For a new purchase with no history, assume your own channel mix in year one, and be honest that most new listings start with one platform doing nearly all the work.

Channel share from the largest platformWhat it means for the model
Under 50%A change at one platform hurts but does not decide the year
50% to 80%A fee or ranking change shows up clearly in annual revenue
Over 80%The platform's policies are a primary driver of the deal
100% with no guest listOne account decision can stop all new revenue

These bands are a way to think, not industry thresholds. The point is to write the number down. Many investors cannot say what share of revenue depends on one company until something goes wrong.

Fee changes: the easiest risk to model

A fee change is the most predictable platform risk, because the platforms publish their fees and their right to change them.

Airbnb's Terms of Service state that "Airbnb reserves the right to change the service fees at any time" and that it "will provide Members notice of any fee changes before they become effective." The same Terms say Airbnb "may modify these Terms at any time," with notice of material changes at least 30 days before they take effect.

Airbnb's service fee page, opened today, describes two structures. Under the split fee, "most hosts pay a 3% service fee, but some pay more," and the guest pays a separate fee. Under the single, host-only fee, "most hosts pay 15.5%, remaining hosts typically pay 14%-16%." The page says this single structure "is mandatory for most hosts, including traditional hospitality listings," that hosts are being moved to it, and that a switch to the single fee cannot be reversed. Our guide to host-only vs. split fees on rental platforms explains how the two compare for guests and hosts.

For underwriting, the lesson is not the exact percentage. It is that the fee line in your model is set by someone else and has changed before. Two habits help:

  • Model the fee you will actually pay, not the one a calculator defaults to. If you will use property management software or are in a market where the host-only fee applies, a 3% host fee is the wrong input.
  • Add a fee-change scenario. Raise the platform fee a few points and see what happens to cash flow and debt coverage. If you move to a host-only fee and raise nightly rates to compensate, model whether your rate still ranks well against comparable homes that did not.

Ranking changes: the hardest risk to see coming

A drop in search ranking cuts bookings without any notice, and you may never be told why.

Airbnb's help page on search says "the quality, popularity, price, and, for homes, location of a listing heavily influence" where it appears. It says "higher quality listings with better ratings and reviews tend to rank higher," that "more popular listings tend to rank higher," and that listings "priced below other comparable listings in the area" tend to rank higher. It also says "one of the ways hosts can most directly affect the performance of their listing in search is by adjusting their price." The Terms describe ranking as depending on "a variety of factors," including guest search parameters, listing characteristics, guest experience, host requirements and guest preferences.

Read those lines as an investor and two things stand out. First, price is a ranking input, so holding a high rate can cost visibility. Second, the factors are broad and the weights are not published, so ranking can change when the platform changes its algorithm, even if nothing about your home changed.

A ranking drop shows up in the model as fewer booked nights, a lower average rate (because you cut price to win back position), or both. Our guide on occupancy rates explains why those two numbers have to be read together.

Several house keys hanging from a single key ring held in a hand
Every key on one ring: when one channel brings every booking, one decision can affect all of them.

Suspension and removal: the tail risk

Suspension or removal is rare for a well-run listing, but it is the scenario that can stop new revenue at once, so it belongs in the model as a tail case.

Airbnb's help page on suspended, restricted and removed listings draws a clear line between them:

  • Suspended listing: "In most cases, you can continue to host your current and upcoming reservations," but the listing is hidden and "you won't be able to receive any new bookings at that listing."
  • Restricted account: a host may be "paid on a different schedule, such as after your guest's checkout date."
  • Removed listing: "Reservations at listings with check-in after the removal date will be cancelled," and the listing "will not be eligible for reactivation unless you successfully appeal the removal."
  • Removed account: "reservations for your listings with check-in dates after the removal date will be canceled."

The same page says removals "are based on repeated or severe violations, rather than a single minor incident," and that hosts may appeal a removal by contacting Airbnb. The Terms (section 12.3) say Airbnb may, "with or without prior notice," "suspend, remove, disable access to, or restrict visibility of Listings." Our guide on what to do if Airbnb suspends your listing walks through the appeal path.

Two things make this a cash flow event and not only a revenue event. A restriction can delay payouts, which strains a property with thin reserves. And an outage in peak season can cost far more than its share of the calendar, because the lost nights are the high-rate ones.

Refunds and rules that move money after the booking

Platform rules also decide how much of a booking you keep, and those losses rarely appear in a pro forma.

Airbnb's Rebooking and Refund Policy says that when a guest is refunded for a reservation issue, "the host will receive no payout or will have their payout reduced by the amount refunded to their guest." Covered issues include a host cancellation, a guest not getting access, a home that is not habitable at check-in, or a home that is significantly different from its listing, such as a missing or broken advertised amenity. Guests "must report Reservation Issues within 72 hours after discovery."

For a model, treat this as a small, steady leak rather than a scenario: a line for refunds and adjustments, informed by the seller's payout history if you can get it. Platform rules on contact also limit how you can move repeat guests off the platform. Our guide to platform rules on off-platform contact explains what the major platforms allow.

How to haircut a projection for platform risk

The cleanest way to price platform risk is a set of named scenarios, each tied to one cause, rather than a single vague discount on revenue.

A blanket "minus 10% for risk" hides what you are protecting against. Separate scenarios tell you which risk the deal is most exposed to and what to do about it. Here is a worked example with invented numbers.

Example scenarioWhat changesQuestion to answer
BaseMarket comps, current feeDoes the deal work at all?
A: fee changePlatform fee up, partial rate recoveryDoes debt coverage hold with a higher fee?
B: ranking dropFewer booked nights, lower rateHow many weak months can reserves carry?
C: suspensionSeveral weeks of no new platform bookings in peak seasonAre reserves enough to pay the loan through it?
D: combinedA and B togetherIs this a deal you would still buy?

Run each one through the same model. Our deal analyzer and the other investor tools let you change one input at a time and see the effect on cash flow and debt coverage. A few rules keep the exercise honest:

  • Use payout, not gross. Debt is paid from what lands in your account.
  • Do not let the haircut drive rates up. If a scenario only works because you assume higher nightly rates, remember that price is a ranking factor.
  • Tie reserves to the worst scenario you accept. If scenario C means six weeks of loan payments with no new bookings, the reserve should cover that, plus the fixed costs that do not stop.
  • Write down the concentration assumption. If you plan to bring the platform's share down over time, state the target and the cost of getting there.

Reducing the risk, and what that costs

You can lower platform risk with more channels and a direct guest list, but each step adds cost and work that a model should include.

StepWhat it reducesWhat it costs
List on a second platformConcentration on one ranking and one accountCalendar sync, a second fee structure, more messages
Direct booking websiteDependence on any platform for repeat guestsWebsite, payments, insurance and marketing
Your own guest listReliance on search for future demandTime, and care to follow each platform's contact rules
Strong operationsRefunds, cancellations and suspension triggersCleaning standards, maintenance, fast responses
Higher reservesThe cash flow hit of any scenarioCapital that is not earning elsewhere

Two honest caveats. More channels are not automatically better: each one adds a sync point where a double booking can happen, and our guide on listing on multiple channels without double bookings covers where that fails. And a direct channel takes years to build; in year one, most of the demand will still come from the platforms. The goal is to reduce concentration over time, not to pretend it away at purchase.

Professional management can help with the operational side, because cancellations, slow responses and listing accuracy drive many platform problems. It does not change the platform's fees or its right to change its rules, and its own fee is one more line to haircut. For a small, simple property with an owner who has the time, self-managing with good systems can carry the same platform risk at lower cost.

Questions to ask before you buy

A short list of questions turns platform risk from a vague worry into a line in the model.

  1. What share of the last twelve months of revenue came from each channel?
  2. What fee structure does the listing use today, and what will it use after the sale?
  3. Has the listing ever been suspended, restricted or delisted, and why?
  4. Do the reviews, ratings and listing history transfer, or will you start a new listing from zero?
  5. How many repeat or direct guests does the property have, and can you contact them within the platforms' rules?
  6. What do the payout reports show for refunds and adjustments?
  7. How many months of loan payments and fixed costs could reserves cover with no new bookings?

Question 4 matters more than many buyers expect. A new listing starts without reviews, and the search page says ratings and reviews affect ranking, so the first months can book below what the seller's history suggests.

The bottom line

Platform risk is a concentration risk, and it belongs in the pro forma as named scenarios: a fee change, a ranking drop and a lost-booking period. Airbnb's own Terms say it can change fees at any time with notice and suspend or remove listings, and its policies say removal cancels future reservations. Measure the channel share, model each scenario on payouts, size reserves to the worst case you accept, and count the cost of diversifying. If you want to talk about investing with JekkyStay, see our investors page.

Frequently asked questions

What is platform risk for a short-term rental investor?

It is the risk that a single booking platform changes something you do not control, such as its fees, its search ranking, its refund rules or your account status, and that the change cuts your revenue. The more of your bookings come from one platform, the larger the risk.

How much can Airbnb change its host fees?

Airbnb's Terms of Service say it reserves the right to change service fees at any time and will give members notice before changes take effect. Its service fee page says most hosts on the single host-only fee pay 15.5%, and that most hosts are being moved to that structure.

What happens to bookings if Airbnb removes a listing?

Airbnb's help page on suspended and removed listings says reservations at listings with check-in after the removal date will be canceled. A suspended listing can usually keep current and upcoming reservations but cannot take new bookings while suspended.

How do I haircut a pro forma for platform risk?

Run separate scenarios rather than one vague discount: a higher platform fee, a lower average rate or fewer booked nights from a ranking drop, and a month or two of lost bookings from a suspension. Check debt coverage and cash reserves in each, and size reserves to the worst one you accept.

Does listing on more channels remove platform risk?

It reduces concentration but does not remove risk. Each channel has its own rules, and more channels add calendar sync, fee and management work. A direct booking channel with your own guest list is the only one whose rules you set.

Can I pay Airbnb for a better search position?

Airbnb's help page on how search works does not describe paid placement for hosts. It says price is one of the ways hosts can most directly affect search performance, and that quality, popularity, price and location heavily influence placement.

Sources

  1. Airbnb Terms of Service
  2. Airbnb Help Center: Airbnb service fees
  3. Airbnb Help Center: How search results work
  4. Airbnb Help Center: Suspended, restricted or removed listings and accounts
  5. Airbnb Help Center: Rebooking and refund policy
  6. Airbnb Help Center: Host cancellation policy

Invest in large group homes

Accredited investors can register for future JekkyStay opportunities.