(Members can access a PDF of this post here).
4Q25 Update.
Airbnb reported 4Q25 and the stock was up 4% in afterhours trading.
Revenues grew +12% y/y to $2.8bn for 4Q, reaccelerating 100bps sequentially.
This marks the second time in the past year that revenues reaccelerated after slowing and puts them above the “at least double digits” revenue goal they set over a year ago.
GBV was even stronger, growing 16% y/y for 4Q (13% FX-adjusted)
This was the strongest growth they had in two years.
Nights and Seats also grew 10% y/y.
On the earnings call Brian Chesky emphasizes that there were “hundreds of improvements the team shipped driving hundreds of millions of dollars in revenue in 2025 alone. And we believe Project Why will deliver hundreds of millions more this year.”
This level of iteration and product improvement is key to their engineering organization and ability to subtly improve customer experience overtime, which shows up in higher booking completion rates.
They also focused this product iteration on localizing for a specific country like Brazil, where they were able to move from a Top 10 Airbnb market to Top 5.
Among one of their newer innovations was “Reserve Now Pay Later”. The ability to book a place without paying has significantly reduced friction. They noted that this increased booking lead time and also led to more larger homes booked.
Another update changed the way their API interacts with PMSs (property management systems). Many hosts cross-list their homes and use software to do that. The software interacts with Airbnb’s API to keep listings from getting double booked. Because of Airbnb’s history of charging the customers most of the commission with an opaque formula, it has been hard for hosts to ensure like for like pricing across platforms. So, they changed the way hosts set prices so it is easier to match it across platforms. This likely had a positive impact on bookings as Airbnb was commonly more expensive as a result of this pricing confusion.
Operating profits for the year were flat at $2.5bn, despite revenues growing 10%. This is because they have increased their product development and marketing costs for Experiences and AI.
However, CEO Brian Chesky is careful to note that they don’t need massive capital investments to take advantage of AI. They are already in the midst of incorporating AI into their app but note that the experience needs to be more compelling than a chatbot.
Right now, text boxes are the most common way users interact with AI, but travel requires more images, so they are reworking how to think about the UI of that interaction. Historically, they have been very strong on UI and so they could potentially come out with a new and novel way for users to book on Airbnb.
The goal is to help users find a listing they will love quicker, while also potentially cross-selling them on Experiences they can book for their trip. Experiences are still in early innings of being launched and they talked about how they are going to focus on city by city to build up listing density, much like how they originally did with Airbnb over a decade ago.
They briefly mentioned how sponsored ads are still a future potential opportunity, but they will have to do it in a way that is native for AI. A loyalty program could also come in the future, but similarly they want to do something different for that. Both have been long-standing opportunities to increase monetization and user loyalty, but Airbnb hasn’t rolled out a product for either yet. In contrast, Booking’s Genius loyalty program has been a strong tool to increase user retention and direct bookings.
The AI Risk.
The big risk investors are now worried about is AI. However, Airbnb is particularly well insulated against AI. Most of their inventory is exclusive to Airbnb with single hosts picking Airbnb for its ease of use. And AI can’t web scrap Airbnb without it’s permission. So, Airbnb already aggregated millions of listings, often with an on the ground effort to do so and replicating that isn’t trivial. AI may be able to book on Airbnb through an API or the “agentic web”, but they won’t be able to easily reach the host otherwise.
Among the inventory that is cross listed to PMS systems, the risk of agents accessing this inventory and circumventing Airbnb is a bit higher. However, users like booking on Airbnb because of the insurance policies and support network. The host also loses insurances if the booking is off platform.
There also is Airbnb’s repository of reviews and rating, which helps weed out lower quality listings and helps hosts feel more comfortable with their guests. AI would need to replicate this and won’t be able to until it has a high level of bookings across millions of properties.
Lastly, what is the interface for this AI? A chatbot seems unlikely to be a great experience so something new will have to be invented. This is exactly what Airbnb is trying to do. As CEO Brian Chesky notes, the AI models are available to everyone to use. So, they will be able to do everything with AI that this theoretical AI agent would be able to.
Now Airbnb does still have some risk here because a very large portion of traffic is direct to them. So, if in the future most travel journeys start with an AI agent, then they may have to pay to acquire that traffic, or the AI agent may be able to book on Airbnb without the user ever going on the site (using the so called “agentic web”). While Airbnb is still a critical piece of infrastructure, providing aggregated listings, review data, insurance, payment facilitation, and support, it does mean that the user will have less loyalty and their margins might be compressed in this sort of transaction.
However, while this is a risk, it doesn’t seem like the most likely of scenarios. When a user wants an Airbnb type home, they think Airbnb and go to Airbnb. It seems unlikely that this behavior will totally change in the future. Users can start these searches today on Google instead of Airbnb, but they pick Airbnb because it is a better experience. With them solely focused on incorporating AI into the travel experience, and other AI companies like Google focusing on bigger opportunities, it makes sense that Airbnb would be able to create a compelling experience before other AI companies.
That’s not to say OpenAI and Google won’t have something for the travel space, but just that it isn’t likely to be such a superior experience that it weighs on Airbnb’s direct traffic. This is especially true because they don’t have access to their inventory as of now.
With technology changing quickly though, there is still risk here.
Valuation.
Airbnb currently has a market cap of $75bn and an enterprise value of $64bn with a gargantuan $11 billion cash pile.
Looking at them on a free cashflow basis though, we need to adjust out their net interest income of ~$600mn to avoid putting a multiple on this low quality stream of earnings (interest on cash). On a trailing EV/ FCF basis they trade at 26x.
However, to get a better sense of returns we turn the Reverse DCF. We show our revenue and margin assumptions below.
Below we can see the outputs at a $122 stock price.
Below we share some call notes.
Call Notes.
Project Y
- Created a small elite team to make it easier to find and book a home on Airbnb.
4 Areas in which they are applying Project Y is Driving Growth
- Pricing
- remove hidden fees and implement price transparency
- flexible cancellation policies
- Reserve Now Pay Later drove bookings in 4Q for larger, high priced homes
- Supply
- Leaned into events (2024 Olympics and 2026 World Cup)
- Removed half a million low quality listings
- Brazil moved from top 10 to top 5 market on Airbnb
- Brazil was the second largest contributor to first time bookers, behind only the U.S.
- Services & Experiences
- Testing grocery delivery and airport pickups
- Hotels
New CTO, Ahmad Al-Dahleto to lead Airbnb AI models (helped build Meta’s Llama)
Regions
- LatAM: high-teens, Brazil saw 20% in nights booked and the number of first-time bookers grew +17% y/y. Origin nights booked in Mexico accelerated to high-teens
- APAC: mid-teens, nights booked in India grew 50% y/y and first-time bookers grew more than 60% y/y
- EMEA: high-single digits
- North America: mid-single digits
4Q Drivers
- Launch of reserve now pay later, updates to their cancellation policy (single service fee), and the beginning of the migration to a simplified fee structure. These three features delivered over 200bps of growth in nights book and ~300bps of growth in GBV
Reserve Now Pay Later
- Drove bookings in 4Q for larger, high-priced homes. Will roll out globally. Has lengthened lead times. Increased aggregate cancellation rate by ~1% from 16% to 17% (higher within cohorts that chooses this product), but is not hugely material relative to the broader cancellations on the platform
Tax Rate
- Starting in 2026, we expect the One Big, Beautiful Bill Act to materially reduce our effective tax rate to the mid-high teens, primarily due to how foreign earnings are taxed, which will benefit consolidated earnings
Share Repurchases
- Repurchased $1.1bn of shares in 4Q
- Used 80% of FCF to repurchase $3.8bn shares in 2025
- Have repurchased 9% of diluted share count since 2022
Guidance
- 1Q26- Expect revenues to grow 14-16% y/y in (includes 3pt FX tailwind). Will see a 50bps incremental revenue in 1Q due to Easter. GBV to increase in low teens, driven by high single digit growth in nights & seats booked and moderate increase in ADR
- FY26- We expect revenue growth to accelerate to low, double digits with an ambition to grow even faster than that
AI
- AI traffic converts better than Google:
- I think these chatbot platforms are going to be very similar to search. They’re going to be really good top of funnel discoveries and in fact, what we’ve seen is I think they’re going to be positive for Airbnb. AI will increase top of funnel awareness for Airbnb just like Google search did. What we see is that traffic that comes from chatbots, converts at a higher rate than traffic that comes from Google. These companies will be very helpful top of funnel traffic generators for Airbnb, just like Google was. AI Search:
- AI Search:
- AI Search is live to a small experiment group. Want to make it more conversational, integrating it into more of the trip, and eventually will be looking at sponsored listings as a result of that
- AI Customer Support Chatbot
- Customer support chatbot is now solving 1/3 of support issues without needing a live specialist, resolution times are faster. Planning to roll out globally
- 2026 AI Expectations if Successful and how it will affect the P&L:
- 1) More than 30% of tickets will be handled by a customer service agent in many more languages and the AI customer service agent will not only be able to chat but will be able to talk. This will decrease the cost base in customer service.
- 2) Make engineers more efficient.
- 3) AI conversational search, not just text
- “By layering AI over the entire Airbnb experience, we believe we’re building something that’s impossible to replicate.”
Repeat Customers
- The repeat rate of Airbnb if you can simplify it down to the satisfaction of the guest. The satisfaction of the guest, first and foremost, is satisfaction of the home, and then if something goes wrong, the satisfaction of customer service. That’s why we focus first and foremost on host quality. The trip quality, which is the score we look at, has gone up significantly. That means satisfaction has gone up. That means repeat use is stronger, and I think that explains much of the reacceleration.
Hotels
- Hotels was a single-digit % of total nights booked, but growing nearly double that of the overall platform. It will take some time for that business to scale to have a meaningful contribution to growth
*At the time of this writing, one or more contributors to this report has a position in ABNB. Furthermore, accounts one or more contributors advise on may also have a position in ABNB. This may change without notice.