Airbnb (ABNB.US) Q3 Conference Call: CEO Weighs Back to Growth Model and Plans to Launch AI Travel Agents in the First Half of Next Year

The Zhitong Finance App learned that after the market on November 5 EST, Airbnb (ABNB.US) held a results conference call after announcing the results for the third quarter of 2026. CEO Brian Chesky announced “Airbnb is back in growth mode” (Airbnb is back in growth mode) at the conference, saying “it is rare for a market platform of our size to accelerate again, but this is exactly what we are doing”. The topics that received the most attention at this conference were the driving factors and sustainability of accelerated growth, and the second was the company's platform strategy in the AI era — Chesky said at the conference that the self-built AI travel agency plans to be launched to tenants in the first half of next year and will open the platform to external AI agents through an MCP agreement. At the same time, he stressed that with the transformation of Airbnb into an AI-native company, “I think the immediate opportunities are even greater”.

In terms of performance, according to LSEG data (CNBC quoted), the company achieved revenue of US$4.87 billion in the third quarter, up 19% year on year (up 16% year on year), higher than the market's consensus forecast of US$4.75 billion, the fastest revenue growth rate in more than three years; diluted earnings per share of US$3.16, better than market expectations of US$2.85; net profit of US$1.89 billion, up 38% year on year, net profit margin of 39%; adjusted EBITDA was US$2.51 billion, up 22% year on year and profit margin 52%. Total bookings (GBV) reached US$27.7 billion, up 21% year on year; night and seat reservations reached 155 million, up 16% year over year. The company directed fourth-quarter revenue of 3.22 billion US dollars to 3.28 billion US dollars (up 16% to 18% year over year), higher than the market's consensus forecast of about US$3.16 billion, raised the annual revenue growth rate guide to high teens, raised the annual adjusted EBITDA margin guide to about 36%, and announced additional stock repurchase authorizations of up to 12 billion US dollars. The financial report was released after the market. On the same day, the stock price closed down 3.5% to 170.13 US dollars during the regular trading session. After the market, it surged about 12% to above 190 US dollars, which exceeded the 7.5% fluctuation of previous pricing in the options market.

Focus and Response: Growth Accelerated Again and AI Agency Strategy

The central focus of this conference is on the sources and sustainability of the renewed acceleration of growth. Financial reports show that revenue, GBV, and overnight growth have all accelerated month-on-month. The five major core markets of the US, France, the United Kingdom, Canada, and Australia have all accelerated. The sharp rise in post-market stock prices also directly reflects the market's acceptance of the “return to growth” narrative. At the conference, Chesky attributed the acceleration to product and operational restructuring over the years: the company released more than 50 products and features this quarter, which is nearly 70% more than a year ago, and AI has significantly accelerated the pace of development. Another hot topic in the Q&A session was the impact of AI agents on platform traffic. In response to concerns about the impact of external general agents, Chesky responded that they received very little traffic and “very few” reservations from general agents. Foreign agents accounted for an negligible share of the platform's nearly 500,000 daily bookings, and saw this as a discovery channel opportunity similar to Google. The overall response of the market was positive, but market opinions also suggest that exceeding expectations in a single quarter was not enough to confirm the sustainability of accelerated growth.

Growth accelerated across the board, and all five core markets accelerated

According to CFO Ellie Mertz, GBV increased 21% year over year in the third quarter, 5 percentage points faster than in the second quarter; overnight and seat reservations increased 16% year over year, with “book now, pay later” (RNPL) contributing a net increase of 3 percentage points; ADR increased 4% year over year (fixed exchange rate caliber increased 5%). Looking at subregions, North America achieved low double-digit growth overnight, the fastest in four years; EMEA grew in mid-teens; Latin America and the Asia-Pacific region grew by more than 20%, and all four major regions accelerated from the second quarter. The growth rate of the five core markets (the US, France, the United Kingdom, Canada, and Australia) all accelerated; in the expansion market, Brazil grew by more than 30%, and India grew by more than 60%. The growth rate of first-time bookers accelerated to 13%, the highest since early 2022.

AI Strategy: Build Your Own Travel Agent and “Post-Dialogue Interface”

Chesky explained in detail the three advantages of self-built AI travel agents: first, the platform has nearly 20 years of market data and knows its own market; second, it can control the reservation itself, including payment, changes, and customer service protection; third, it can provide an exclusive travel interface — combining components such as maps, photos, comparisons, calendars, and itineraries in real time through a “generative UI” (generative UI). He called it a “post-chat interface” (post-chat interface), and clearly opposed the general agent form of plain text chat. The product is scheduled to be rolled out to guests in the first half of next year. In terms of AI application implementation, AI customer service assistants already support more than 50 languages, about 40% of customer service issues can be solved without human intervention (about one-third in the second quarter), and has been expanded to US voice customer service. Customer service NPS reached a record high; customer service costs for a single reservation fell by more than 17% year on year.

Hospitality and Local Services Expansion: Targeting the $900 billion market

Chesky said that the hotel business started with a pilot in 4 cities, and now there are thousands of boutique and independent hotels in more than 60 destinations, and hotel reservations accelerated again in the third quarter. He said that hotels are a 900 billion dollar market, and the next few years will bring “multiple billion of dollars in revenue” (multi-billion dollar revenue) to the company, and admits that Airbnb's early slogan was “Forget Hotels” (forget hotels), but now the attitude has changed 180 degrees. Approximately 35% of first-time hotel guests will return to book a homestay within a year. In September, the company hired former Booking.com core executive Pepijn Rijvers as Chief Business Officer (Chief Business Officer) to oversee the real estate and hotel business and international expansion. In terms of service, car rental launched in summer is one of the fastest growing services. Local services such as food delivery, laundry, baby products, ski equipment, and boat rental were also announced in the third quarter.

Pricing, RNPL and monetization rate: Q4 will usher in a reversal

Mertz explained that the implied monetization rate (take rate) for the third quarter was 17.6%, down slightly year-on-year, mainly due to RNPL expansion boosting GBV denominators and incentives and deductions for new businesses, which were partially offset by travel insurance and single service fees; as the RNPL denominator effect subsides, the fourth quarter monetization rate is expected to rise slightly year-on-year. The RNPL was launched in the US a year ago and expanded to global flexible and moderate cancellation policy listings at the beginning of the third quarter. It has now covered more than 40% of GBV. Q3 is the peak of RNPL's net impact on overnight stays, and the net increase in the fourth quarter will decline sequentially; along with the expansion, the platform cancellation rate rose to about 19% (about 16% last year). The company completed the simplification of service fees in October, and almost all listings switched to a single service fee of 15.5%. Chesky added that pricing is one of the company's biggest levers. Dynamic pricing can “cut prices by $10 and significantly increase occupancy rates” while making landlords earn more. Furthermore, travel insurance revenue increased nearly 70% year over year and currently accounts for a low single-digit percentage of bookings.

Social graph, exclusive inventory, and loyalty program

Looking back on the golden age of social networking from 2004 to 2008, Chesky said that the real social graph disappeared in the evolution of social media, and that the travel map that Airbnb launched in May this year is reconstructing the real-world social map. He called it “one of the underrated parts of the business,” and that more social features will be introduced next year. In terms of inventory, more than 60% of the platform's listings are exclusive inventory. He believes that the exclusive ratio is not important; only the absolute quantity matters. The company is also developing sponsored listings (sponsored listings) ads, but they will be designed in the context of AI agents. Regarding loyalty programs, he said “we are absolutely looking at a loyalty program” (we're definitely looking at a loyalty program), pointing out that Airbnb is the only major travel company without a loyalty program and is at a disadvantage. The more vertical categories, the more attractive the program.

Capital return and guidance: additional $12 billion in repurchases

The company repurchased 700 million US dollars of Class A shares in the third quarter, with a remaining authorization of 2.7 billion US dollars, and announced an additional repurchase authorization of up to 12 billion US dollars. In terms of free cash flow, the third quarter was 1.4 billion US dollars, with a cumulative total of 4.9 billion US dollars over the past 12 months. The FCF profit margin was 35%, affected by RNPL's working capital occupation and the capital expenditure of the new New York office building of 75 million US dollars. In terms of guidance, fourth-quarter revenue is expected to be between US$3.22 billion and US$3.28 billion, up 16% to 18% year on year (including a favorable exchange rate of about 2 percentage points); GBV is growing in mid-teens, with a low double digit increase overnight; and adjusted EBITDA profit margins decline year on year (partly due to the pace of marketing investment). The full-year 2026 revenue growth guideline was raised from at least mid-teens in the previous quarter to high teens, and the adjusted EBITDA profit margin guide was raised from at least 35.5% to about 36%. Mertz suggested that the company is closely monitoring the situation in Europe (France in particular) and the Middle East conflict, but the guidelines do not assume a significant continuing impact.

Full Meeting Minutes

Andrew Slabin (Vice President of Investor Relations): Welcome to Airbnb's 2026 Q3 Earnings Conference Call. Co-Founder and CEO Brian Chesky and Chief Financial Officer Ellie Mertz attended the conference. Shareholders' letters have been posted on the company's investor relations website. Before I begin, I would like to remind everyone that the statements at this meeting include forward-looking statements, and actual results may differ materially from the forward-looking statements; non-GAAP financial indicators will be covered in the meeting, and reconciliation statements with GAAP indicators are provided in the company's financial statements.

Brian Chesky (Co-founder and CEO): Airbnb is back in growth mode (“Airbnb is back in growth mode”). It's rare for a market platform of our size to re-accelerate growth, but that's exactly what we're doing. Many years ago, we began rebuilding our technology platform and operating methods, using a single roadmap (single roadmap) and a model focused on small teams. Inspired by the early “Project Hawaii” project, hundreds of product improvements have accumulated hundreds of millions of dollars in revenue; the work of the pricing team also brought in hundreds of millions of dollars in incremental revenue. In the third quarter, we released more than 50 products and features, which is nearly 70% more than a year ago. AI significantly accelerated the pace of development.

Product side: After the launch of Neighborhood Pages (Community Pages), 75% of guests use Airbnb maps; the Paris pilot brought significant booking growth and will cover more than 1,500 communities around the world by the end of the year. In terms of social functionality, we have expanded Connections and launched new travel maps (travel maps). The host side launched a multi-listing calendar (multi-listing calendar), and nearly a quarter of eligible hosts joined in the first month of opening.

Personnel aspects: Pepijn Rijvers joined in September as Chief Business Officer (Chief Business Officer). He was previously the core leader of Booking.com, managing its entire lodging business and largest performance marketing team, and now oversees our listing and hotel business and international expansion.

On the hotel side: We started with a pilot in 4 cities, and now we have thousands of boutique and independent hotels in more than 60 destinations, and hotel reservations accelerated again in the third quarter. Approximately 35% of first-time hotel guests will return to book a homestay within a year. In terms of service, car rental launched in summer is one of our fastest-growing services; in the third quarter, we also announced local services such as food delivery, laundry, baby products, ski equipment, and boat rental.

Cooperation aspects: We have reached cooperation with Delta, Qantas (Qantas), and TAP Air Portugal (TAP Air Portugal); in October, we launched the biggest partnership in the company's history — an exclusive partnership with American Express (American Express), and more than 1 million homes entered Amex Travel. There are currently thousands of bookings, with nearly a quarter of them coming from new users, and Luxe's share of listings is higher than normal on the platform.

AI aspect: We have launched AI search, which supports natural language input. It is at the top of apps and websites, and is very popular with Gen Z. It can recommend filters, summarize listings, and compare collections. The landlord side launched AI to generate 3D floor plans. The AI customer service assistant supports more than 50 languages, and about 40% of customer service issues can be solved without human intervention (about one-third last quarter). It has now expanded to US voice customer service, and customer service NPS has reached a record high.

Regarding self-built AI travel agents: We think there are three major advantages. First, we know our own market best and have nearly 20 years of data; second, we master the reservation itself, including payment, changes, and customer service protection; third, we have an exclusive travel interface — combining maps, photos, comparisons, calendars, itineraries, etc. in real time through a “generative UI” (generative UI). This is a “post-chat interface” (post-chat interface) rather than a plain text chat method. We plan to launch this product to guests in the first half of next year.

Ellie Mertz (Chief Financial Officer): GBV for the third quarter was US$27.7 billion, up 21% year on year, 5 percentage points faster than in the second quarter; overnight and seat reservations were 155 million, up 16% year over year, with RNPL contributing a net increase of 3 percentage points; ADR increased 4% year over year (excluding exchange rate impact of 5%). Revenue was US$4.87 billion, up 19% year on year (fixed exchange rate caliber increased 16%), the fastest growth rate in more than three years.

Looking at the subregion, overnight growth in North America was low in double digits, with the fastest increase in four years; the EMEA grew mid-teens, and the UK, France, Germany, Spain, and Italy accelerated month-on-month across the board; Latin America and the Asia-Pacific region grew by more than 20%, with Brazil continuing to grow by more than 30%, and India growing by more than 60%. The growth rate of the five core markets (the US, France, the United Kingdom, Canada, and Australia) all accelerated.

RNPL timeline: Launched in the US a year ago, expanded to properties with flexible and moderate cancellation policies around the world at the beginning of the year, and expanded to listings with limited and strict cancellation policies at the beginning of the third quarter. Currently, RNPL covers over 40% GBV. As coverage expanded, the platform cancellation rate rose to around 19% (about 16% in the same period last year). We completed service fee simplification in October, and almost all listings switched to a single 15.5% service fee, which improved the competitiveness and transparency of landlords in pricing.

The implied monetization rate for the third quarter was 17.6%, down slightly year-on-year, mainly due to RNPL expansion boosting the GBV denominator and incentives and deductions for new businesses, partly offset by travel insurance and single service fees; the fourth quarter's monetization rate is expected to rise slightly year-on-year, and the single service fee and travel insurance will bring about a moderate increase in the long term.

Profit side: Net profit of US$1.9 billion, net interest rate of 39%, benefiting from a reduction of US$135 million in income tax provisions; the results for the same period last year included a deferred income tax valuation provision of US$213 million. Adjusted EBITDA was US$2.5 billion, up 22% year over year, with a profit margin of 52% (50% in the same period last year). In terms of free cash flow, the third quarter was 1.4 billion US dollars, with a cumulative total of 4.9 billion US dollars over the past 12 months. The FCF profit margin was 35%, affected by RNPL's working capital occupation and the capital expenditure of the new New York office building of 75 million US dollars.

In terms of return on capital: $700 million of Class A shares were repurchased this quarter, with a remaining authorization of 2.7 billion US dollars; the company approved a new repurchase authorization of up to 12 billion US dollars.

In terms of users and efficiency: The growth of first-time bookers accelerated to 13%, the highest since the beginning of 2022; the number of in-app nights increased 29% year-on-year, accounting for 66% of the total number of nights. AI customer service solved about 40% of the problems, and customer service costs for a single reservation fell by more than 17% year over year. Travel insurance revenue increased nearly 70% year over year.

Guidance: Fourth-quarter revenue is expected to be US$3.22 billion to US$3.28 billion, up 16% to 18% year on year, including a favorable exchange rate of about 2 percentage points; GBV growth in mid-teens; low double-digit growth overnight; adjusted EBITDA profit margin slightly lower than the same period last year, partly affected by the pace of marketing investment. The full-year 2026 revenue growth guideline was raised to high teens (the previous quarter's guideline was at least mid-teens), and the adjusted EBITDA profit margin guideline was about 36% (at least 35.5% in the previous quarter). We are closely monitoring the situation in Europe (France in particular) and the Middle East conflict, but the guidelines do not assume a significant continuing impact; global travel demand is still healthy at the bottom of the world.

Q & A session

Richard Clark (Bernstein Research): Please talk about the progress of dynamic pricing and the fourth quarter monetization rate outlook?

Brian Chesky: Pricing is one of our biggest levers. In Pepijn's words, “if you think reserve now, pay later is big, wait till you see what we can do with pricing” (if you think the scale of booking first is huge, wait and see what we can do in terms of pricing). Customers only care about two things: product and price, each accounting for 50%. Dynamic pricing can simultaneously lower prices and make landlords more money — for example, by cutting the price by $10, the occupancy rate will increase dramatically. The adoption rate is very good, but no specific data will be disclosed for the time being; AI is very good at optimizing pricing, and will see results in the next few quarters.

Ellie Mertz: The RNPL in the third quarter boosted the GBV denominator and lowered the monetization rate. This effect will reverse in the fourth quarter; a single service fee and travel insurance will bring about a moderate increase in the long term.

Lloyd Walmsley (Mizuho Research Division): How large-scale can the hotel business be? Also, how can AI agents “get out” of Airbnb's walls?

Brian Chesky: The growth rate of hotels is significantly higher than that of the platform as a whole. Hotels are a $900 billion market, and the next few years will bring us billions of dollars (“multiple billion of dollars”) in revenue. We don't need to win the entire market; we just need people staying in hotels to place orders on Airbnb. Our slogan in the early years was “Forget Hotels” (Forget Hotels), but now our attitude has changed 180 degrees. The hospitality industry welcomes Airbnb — we have a Gen Z, high-end, and global audience. We will do performance marketing with high efficiency. Regarding proxy opening: We will break out of Airbnb's walls through MCP (Model Context Protocol) integration and integrate multiple MCPs in the first half of next year. We don't build generic agents - the generic interface is only full text, while dedicated domain agents can generate customized UI, such as a car rental interface. This is the “post-chat interface” (post-chat interface).

Ken Gawrelski (Wells Fargo Research): What do you think of the dispute between vertical agents and generic agents? The impact of the RNPL expansion on cancellation rates?

Brian Chesky: We'll be integrating with a large number of vendors, and every aspect of the trip will have MCP and API. The ultimate goal is that all vertical travel categories can be booked natively within the app — there is currently no schedule. Rental cars are now link-off through “Trawler” partners, and native integration will gradually be implemented. I don't think there will be a scenario where “one chatbot dominates everything like Google” - the operating system level won't allow this, such as Apple. Generic agents are suitable for simple everyday tasks, but specialized agents win out in complex tasks; we can obtain the same data by connecting to Gmail, Google Calendar, etc. through MCP. Our standard is: it must be 2 times, 3 times, 5 times better, not 20%.

Ellie Mertz: The pace of RNPL expansion is as described above; the third quarter was the peak of RNPL's net impact on the night, and the net increase in the fourth quarter will decline month-on-month.

Justin Post (Bank of America Research): What is the exclusive inventory strategy? How much room is left for insurance penetration?

Brian Chesky: Over 60% of our inventory is exclusive to Airbnb. But our philosophy is: exclusive proportions don't matter; absolute quantity matters. For example, out of two stores, if one store has 10% more exclusive products, the increase in traffic may increase by more than 10%. We want to have exclusivity in every category — masseuses, chefs, photographers, exclusive experiences; at the same time, use major products such as hotels and car rentals to become a one-stop shop.

Ellie Mertz: Travel insurance currently accounts for a low single-digit percentage of bookings, a sharp increase over the previous year, and the profit margin is very good. We're considering introducing more insurance products for renters and landlords.

Benjamin Black (Deutsche Bank Research Division): What is the long-term vision for social discovery? How to consider sponsored listing ads? Is there any progress with the loyalty program?

Brian Chesky: 2004 to 2008 was the golden age of social networking, then evolved into social media, and the real social graph disappeared. Airbnb launched a travel map in May this year and is reconstructing a real-world social map — your travel companions and people you meet during your trip. I think this is one of the underrated parts of the business, and more social features will be introduced next year. Social networking is an agent indicator of trust and will be incorporated into search results; it will make group trips more collaborative, easier to recommend, and a more sticky platform. Gen Z is our fastest-growing customer group. They have two major demands: saving money (cost of living, student loans, AI squeezing entry-level jobs) and socializing; the same goes for the “empty nest” group on the other end. Regarding the sponsorship list: We are developing it, but we will design it in the context of an AI agent — launch an agency first and then determine the advertising model. About loyalty programs: “We are absolutely looking at a loyalty program” (We're definitely looking at loyalty programs). Apple doesn't have a loyalty program but the products are loved, but Airbnb is the only major travel company that doesn't have a loyalty program, putting us at a disadvantage; the more vertical we have, the more attractive loyalty programs are.

Jed Kelly (Oppenheimer Research): How do you view the balance between growth and profit?

Ellie Mertz: Over the past two years, we've clearly prioritized growth. The results proved we were able to deliver growth while still being able to continue to reinvest. We're planning for 2027, and continuing the growth momentum remains a top priority. This business is extremely profitable (“extremely profitable”), has strong cash flow, and plenty of room to invest.

Matthew Cost (Morgan Stanley Research): What is the user base for hotel reservations? Will external AI agents divert platform traffic?

Brian Chesky: I saw very little traffic and “very few” bookings from the general agent Instinct. Out of our nearly 500,000 daily bookings, the share of foreign agents is negligible. After peaking in early October, relevant traffic has remained stable for four weeks. We see it as an opportunity rather than a threat — it's a discovery channel like Google. You can use “single product type and project retailer” as an analogy: Airbnb is a project/full-process retailer. 90% of the guests on the platform will contact the host, and the host wants to know the guest rather than the agent. Our interface is extremely visual — each listing has an average of 25 photos; it's highly collaborative; the trip itself is an expected experience — 80% of travelers enjoy the process of researching trips, and most of the traffic occurs during work hours — people are “secretly using Airbnb.”

Ellie Mertz: The ratio of new and existing users in hotel bookings is consistent with the overall platform. The hotel unlocked new usage scenarios for old users, and also provided an entrance for new users with familiar categories.

Concluding remarks

Brian Chesky: Airbnb is back in growth mode and will continue (“Airbnb is back in growth mode and is here to stay”). As we make Airbnb an AI-native company, I think the opportunity ahead of us is even bigger (“As we build Airbnb to become an AI-native company, I actually think the opportunity ahead of us is even bigger”). Finally, I would like to thank the team - “it's all about the people” (“it's all about the people”), team execution is the core reason for growth, and they are really starting to get into the flow (“they're really starting to get into flow”). We have only just begun to return to growth mode (“We are just getting started back in growth mode”).