爱彼迎 (ABNB) 2026年第三季度业绩电话会：增长加速，上调业绩指引

核心要点

• 爱彼迎2026年第三季度营收同比增长19%至49亿美元，创三年多以来最快增速；总预订金额（GBV）同比增长21%至277亿美元。
• 预订间夜及席位数增长16%至1.55亿。其中“先订后付”服务贡献了3个百分点的净提振，剔除该影响后的增幅约为13%。
• 得益于营收增长、成本管控以及AI驱动的客户支持效率提升，调整后EBITDA同比增长22%至25亿美元，调整后EBITDA利润率从50%扩大至52%。
• 管理层上调了2026年全年业绩指引，预计营收增速将达到10%高段，调整后EBITDA利润率约为36%。
• 爱彼迎正将业务从民宿拓展至酒店和服务领域。第三季度酒店预订量提速，管理层表示酒店业务未来几年有望带来数十亿美元的年收入，但未给出具体时间表。
• AI在无人工介入的情况下解决了约40%的客户支持问题，高于第二季度的三分之一，助力单笔预订的客户支持成本同比下降逾17%。

关键财务数据

此外，爱彼迎宣布在剩余授权额度之外，新增一项120亿美元的股票回购计划。

业务与运营表现

爱彼迎在各区域市场的业务增长均出现提速。北美地区预订间夜数录得低双位数增长，欧洲、中东和非洲地区（EMEA）录得10%中段增长，拉丁美洲和亚太地区的增幅均超过20%。其中北美间夜数创四年来最快增速。

拓展市场表现同样强劲。此类市场的房客预订间夜数创近三年来最快增速，其中巴西增长逾30%，印度增长逾60%。

首次预订用户增速提升至13%，创2022年初以来最高水平，其中Z世代是增长最快的客群。通过爱彼迎App预订的间夜数增长25%，该App占总预订间夜数的66%。

第三季度，“先订后付”已覆盖平台总预订金额的40%以上。该产品拉动了预订量，延长了提前预订窗口期，并带动了单价更高房源的预订。整体退订率上升了200至300个基点，从上年同期的约16%升至约19%。

酒店业务依然是战略性增长领域。爱彼迎在60多个主要目的地拥有数千家精品及独立酒店，且本季度酒店预订量有所提速。公司表示，首次预订为酒店的房客中，约有35%会在一年内重返平台预订民宿。

爱彼迎还拓展了多项服务，包括汽车租赁、外卖送餐、洗衣、婴儿用品以及滑雪用具和游艇租赁等本地特色项目。与美国运通的合作将超过100万套爱彼迎房源引入运通旅行（Amex Travel）。通过该渠道产生的前期预订中，有近四分之一来自首次使用爱彼迎的新房客。

AI加快了爱彼迎的产品开发节奏。公司在第三季度推出了50多款产品和功能，较上年同期增加近70%。新增功能包括自然语言AI搜索、房源摘要、比对工具以及房东工具。

公司计划于明年上半年开始推出专注于旅游场景的AI智能体。管理层表示，该交互界面将整合地图、照片、日历、比对和行程规划功能，而不仅仅依赖纯文本聊天。

管理层业绩指引

针对2026年第四季度，管理层预计：

• 营收为32.2亿至32.8亿美元，对应同比增长16%至18%。
• 对冲后汇率将带来约2个百分点的顺风提振。
• 总预订金额增幅在10%中段。
• 预订间夜及席位数至少实现低双位数增长。
• 日均房价（ADR）小幅上涨。
• 隐含变现率同比小幅上升。
• 调整后EBITDA实现同比增长，但受营销支出时间差等因素影响，调整后EBITDA利润率将略低于2025年第四季度。

2026年全年方面，爱彼迎将营收增速预期从此前的百分之十几上调至10%高段；同时将调整后EBITDA利润率指引从先前的至少35.5%上调至约36%。

管理层预计全年调整后EBITDA利润率将略高于自由现金流利润率，原因是“先订后付”推迟了客户付款，从而影响了营运资金的时间节点。

风险与关注重点

“先订后付”虽支撑了预订量增长，但也导致退订增加并推迟了房客付款，对营运资金和短期自由现金流造成压力。管理层指出，第三季度对预订间夜数的净拉动已达峰值，随着退订影响全面显现，预计第四季度的贡献将环比下降。

第三季度隐含变现率略有下滑，原因是“先订后付”在确认相关收入前推高了总预订金额。此外，新业务举措的激励和抵用金也带来了压力，不过部分被旅行保险以及转向单一服务费模式所抵消。

爱彼迎的第四季度业绩展望未假定欧洲局势（包括法国）或中东冲突会带来实质性、持续性的影响。管理层表示正在密切监测这两个地区。

AI旅游智能体仍处于开发阶段。管理层坦言，专用智能体必须大幅优于通用智能体，才能克服用户的使用惯性与转换成本。

分析师问答环节要点

动态定价：管理层将定价机制视为爱彼迎最大的增长杠杆之一。动态定价在初期的应用表现强劲，公司旨在通过提高入住率来增加房东收入，同时降低房客在特定间夜的预订成本。

酒店业务机遇：管理层表示，尽管基数较小，但酒店业务的增速明显快于爱彼迎的核心业务。现有的自然流量可为酒店业务的增长提供支撑，因为未能在平台上找到合适民宿的用户原本可能会离开爱彼迎转去其他平台预订酒店。

AI与外部旅游库存：爱彼迎计划通过API和模型上下文协议（Model Context Protocol）接入更多旅游供应商。长远目标是在爱彼迎App内实现各旅游品类的原生预订，不过管理层尚未提供具体时间表。

保险业务：旅行保险收入同比增长近70%，各地区的附购率均有所提升。目前购买保险的预订比例仍处于个位数百分比，管理层强调了该业务可观的利润率，以及未来推出更多针对房客和房东产品的潜力。

房源差异化：爱彼迎估计其超过60%的房源为平台独有。管理层表示，随着爱彼迎引入酒店和租车等标准化品类，独家房源的绝对数量比其所占比例更为重要。

忠诚度计划与社交功能：管理层证实，随着爱彼迎增加更多旅游品类，公司正在评估推出会员忠诚度计划。此外，公司还计划拓展社交旅行功能，旨在增强信任度、支持多人协作行程规划以及好友推荐。

财报电话会议完整实录

---

完整财报电话会议逐字稿

管理层陈述

Operator

Good afternoon, and thank you for joining Airbnb's Earnings Conference Call for the Third Quarter of 2026. As a reminder, this conference call is being recorded and will be available for replay from the Investor Relations section of Airbnb's website following this call.

I will now hand the call over to Andrew Slabin, Vice President of Investor Relations. Please go ahead.

Andrew Slabin

Good afternoon, and welcome to Airbnb's Third Quarter of 2026 Earnings Call. Thank you for joining us today. On the call with us, we have Airbnb's Co-Founder and CEO, Brian Chesky, and our Chief Financial Officer, Ellie Mertz. Earlier today, we issued a shareholder letter with our financial results and commentary for our third quarter of 2026. These items were also posted on the Investor Relations section of Airbnb's website. During the call, we will make brief opening remarks and then spend the remainder of time on Q&A.

Before I turn it over to Brian, I'd like to remind everyone that we will be making forward-looking statements on this call that involve a number of risks and uncertainties. Actual results may differ materially from those expressed or implied in the forward-looking statements due to a variety of factors. These factors are described under forward-looking statements in our shareholder letter and in our most recent filings with the Securities and Exchange Commission. That being said, we urge you to consider these factors and remind you that we undertake no obligation to update the information contained on this call to reflect subsequent events or circumstances.

You should be aware that these statements could be considered estimates only and are not a guarantee of future performance. Also during this call, we will discuss some non-GAAP financial measures. We've provided reconciliations to the most directly comparable GAAP financial measures in the shareholder letter posted to our Investor Relations website. These non-GAAP measures are not intended to be a substitute for our GAAP results.

And with that, I will turn the call over to Brian.

Brian Chesky

All right. Hey, everyone. Thanks for joining. When you look back at our results, over the last several quarters, there is a clear story emerging. Airbnb is back in growth mode. It's rare for a marketplace of our size to reaccelerate growth, but that's exactly what we're doing. Q3 was our fastest revenue growth in more than 3 years. Revenue grew 19% year-over-year to $4.9 billion. Gross booking value grew 21% to $27.7 billion and Nights and Seats Booked grew 16% to $155 million.

We've also exceeded our outlook across every key metric, and we're seeing the acceleration across the business. In Q3, every region developed to deliver double-digit nights growth. Growth accelerated in all 5 of our core markets: U.S., France, U.K., Canada and Australia, nearly all of our expansion markets. In North America, nights booked grew at our fastest pace in 4 years.

We're also seeing more guests return to Airbnb and more people try it for the first time. First-time booker growth accelerated 13%, now this is the highest it's been since early 2022, and the fastest-growing segment is Gen Z. Our nights booked in our app grew 25% and now 66% of our total nights booked are in our app.

Now I want to start by explaining what's driving this growth. This did not happen overnight. Airbnb grew incredibly fast when we started. But eventually, our technology and the way we operate couldn't keep up -- so a number of years ago, we started by rebuilding the foundation. This was immediately after our IPO. We spent years rebuilding our technology platform. We redesigned the app. We improved our core experience. And a lot of this was pretty standless at the time because we had to trade in some near-term growth to make these very important investments. We did this for the long-term growth of the company. And technology alone was not enough.

We also had to fundamentally change how the company operates. We organized the met around a single road map, we brought all these teams together, and we focused small teams on very specific problems. Then we developed an entirely new way to build products and it was inspired by the original early days of Airbnb of my Co-Founder. And Project Hawaii, this -- what I've talked about in prior earnings calls was our first big bid point. I took a very small team I worked with them very nearly daily, we had 1 job, make it easier to find and book a home. And this team, they work relentlessly, and they ship hundreds of improvements that generated hundreds of millions of dollars of revenue. And so we're like, wow, that works.

Let's take this approach and we start applying it to other teams. So we took this and we applied it to the pricing team, and they were able to deliver hundreds of millions of dollars in incremental revenue. then we took it to international expansion. We gave them more autonomy moved really quickly, and we're seeing results of that team as well. For example, in Q3, nice booked by guests in our expansion markets grew at their fastest pace in nearly 3 years. Brazil grew more than 30% and India grew more than 60%. We also began building new businesses, including hotels and services, and we broadened experienced leaders to help scale them.

And then AI change the speed of everything. We spent years laying the groundwork, but now AI is helping us build, test and improve products faster than we've ever done before. In Q3 alone, we shipped more than 50 products and features. -- nearly 70% more than a year ago because of AI. So when you put it all together, that's what's driving our growth. We spent years changing how urban be operates. And now we're seeing the results.

Now let me share a few updates in the quarter, starting with our fall update. When people book a trip, they don't just care about the home, they care about where it's located. In fact, 75% of our guests use every maps in deciding where to stay. That's why we introduced Neighborhood pages, which help guests get a better feel for an area before booking, including the best restaurant activities and more. Now we started these Neighborhoods pages in Paris, where we saw a really big meaningful increase in bookings. So now what we're doing is we're bringing the neighbor of pages to over 1,500 Neighborhood pages around the world on Airbnb by the end of the year.

Now this is going to be massive because when you stay in Airbnb, you're often staying in a neighborhood and how do you get that information, it's really hard to find in Google. And this is going to be really useful. We're also helping guests plan trips with recommendations from friends and family. We expanded connection, so guests can see where people they know stayed and what they recommend. And our new travel map brings those trips and recommendation together in 1 place.

Now for host, we designed our multi-listing calendar. It makes it easy to manage multiple listings, update pricing and fill up at night. And the first month of early access, nearly 1 in 4 eligible hosts opted in. We're also bringing in rural cost leaders to help drive our next phase of growth. In September, Pine Rivers joined us as Chief Business Officer. Pine was one of the key leaders behind Booking.com's growth. He ran its entire accommodations business, including hotels and homes and led one of the largest for marketing operations in the world. He now leads Airbus Homes and hotels businesses as well as our international expansion.

Now one of the biggest opportunities ahead of us is hotels. And we started small with a pilot in just 4 cities. Today, we have thousands of boutique and invented hotels in Airbnb across more than 60 top destinations, and hotel bookings accelerated again in Q3. Now what's even more exciting is that hotels are bringing new guests to Airbnb. In fact, approximately 35% of guests reversed whose first booking is a hotel. They come back and they book a home within a year. The hotels aren't just a new business for Airbnb. They're also helping us grow our homes business. Services are also becoming a bigger part of Airbnb. We launched car rentals this summer, and they've quickly become one of our fastest growing services.

In Q3, we announced new services, including meal delivery, laundry, baby gear, and we also have local offerings like ski gear if you go in skiing or boat rentals, if you near an ocean or lake. And one of our biggest untapped growth opportunities is actually partnerships. Because for nearly 18 years, partnerships really weren't a major part of how we grew. It was mostly just organic. That's changing because we partner with airlines like Delta, Qantas and TAP Air to reach millings travelers through companies they already use.

In October, we launched our biggest partnership ever, an exclusive partnership with American Express, bringing more than 1 million Army listings to Amex Travel. Now it's still early, but we're seeing thousands of bookings through Amex Travel with nearly 1/4 of them coming from guests who are new to Airbnb and we're seeing share of luxe booking and we typically see on Airbnb. So of course, they're bringing in a really high-end segment of travelers that we haven't typically had as much access to.

Now I want to spend a few minutes talking about AI. I believe AI is one of the best things to ever happen Airbnb. And I want to give you a few examples of what we're doing with it. In Q3, we brought AI into more of the guest and host experience. For guests, we introduced AI search, so they can describe we're looking for in their own words and find the right home service experiences for the trip. So if you want a CII search, just go to our app or a website and you see the very top of the page, there's a toggle. If you open the turn toggle on you can basically type entire sentences like natural language, and then it will basically select all the filters. It's working really well. It's especially popular with Gen Z and younger travelers are very AI native.

AI can also suggest relevant social search filters I can summarize important details of homes and help guests compare places their savings. For host, we're using AI to make it easier to create 3D 4 plans. These are really cool. They can help you price your homes competitively and understand how the listings are performing.

Now we're all seeing AI to transform customer support. Our AI system is now available in more than 50 languages, and this next stat is really key. In Q3, -- approximately 40% of all customer support issues were resolved without a human agent. And this is up from 1/3 in Q2. So this has been huge for us. and we're expanding our assistant to voice support in the U.S., so guests and hosts can help get faster help over the phone. What this means is that customer support cost per booking has declined more than 17% and year-over-year, driven in large part by improvements to our AI assistant.

And if you're wondering, customer service NPS is as high as it's ever been. But these improvements are just the beginning. There's a lot of discussion about whether people will eventually use one AI agent to do everything. And if they do, what does that mean for Airbnb? Are we going to be put out? Well, I don't think people are going to use one AI agent for everything. You see today's AI agents like Muse and Instinct are good at simple routine tasks like making a restaurant reservation or scheduling an appointment. But planning a trip is different. Travel is aspirational, discovering where to go and where to stay as part of the experience.

In fact, 80% of travelers say they enjoy spending time researching trips. People want to look at photos. They want to compare home. They want to make this the people they're traveling with -- and I don't think the general purpose agents are the best to do that. So that is why we are building our own AI agent. And we have 3 distinct advantages when we build our own agent. First, we understand the Airbnb marketplace better than anyone. For nearly 20 years, we've helped guest or host and host find places to stay with -- we help guests find places to stay with Molina host. Every home is different. -- guest care about the home, the location, the reviews and the person hosting them. And we spend years learning how guests make the decision and how us operate. So the understanding gives us a significant advantage in building the AI agent help people find the book the right place to stay. Second, we are part of the reservation itself.

Now this is important because we know that travel doesn't end with the booking. In fact, that's when it starts. Because the everybody handles reservations, payments to customer support, our agent is able to help guests throughout their trip. This means we can manage reservations. We can help you make changes to your plans and accessing Airbnb support and protection when they need them. These are things in outstate agents cannot do easily on their own.

And third, and perhaps most importantly, we're building a new interface specifically for travel. We don't think the traditional chatbot is the right way to plan travel. It's the wrong interface. So what we are doing is developing a custom-designed, AI native experience to respond to what guests are looking for. So instead of relying on traditional chat alone, which is very tech heavy, which again is definitely not the way you should plan travel, -- our agent draws from a library of hundreds of travel-specific interface components, maps, photos, comparisons, calendars, itineraries and compose them in real time to match what they're looking for.

This is going to feel a little bit like generative UI in the sense that it's going to offer up the right interface at the right time for your specific travel needs. Now it's going to help guests explore options, make decisions with people they're traveling with and take action trucking. And over time, we want our agent to help guests plan their entire trip on airing all in 1 place. We're going to start rolling it out to guests in the first half of next year. So I'm going to end where I started. Airbnb is back in growth mode. We're excited about what we're building and with the momentum we're seeing across the business, we're raising our full year outlook.

With that, I'm going to turn it over to Ellie.

Ellie Mertz

Thanks, Brian, and good afternoon, everyone. I'll start with Q3 results and then to our outlook for Q4 and the full year 2026. As Brian shared, Q3 was a strong quarter with a meaningful acceleration across the business. Gross booking value was nearly $28 billion, up 21% year-over-year, a 5 percentage point acceleration from Q2. This was driven by strong growth in Nights and Seats booked and a moderate increase in ADR. Nights and Seats booked grew 16% year-over-year, accelerating significantly for Q2, with strength across all major regions and market types.

Q3 nights benefited from a 3% net lift from the expansion of Reserve Now Pay Later. So excluding the impact of RPL, Nights and Seats booked growth was approximately 13%, still accelerating from Q2. ADR was up 4% year-over-year or 5% excluding FX and who was up across all regions with notable strength in North America and Europe. On a regional basis, we saw low double-digit growth in North America, mid-teens growth in EMEA, over 20% growth in both Latin America and Asia Pacific. These results were an acceleration across all 4 regions relative to Q2.

Turning to our Q3 financial results. Revenue grew 19% year-over-year to $4.9 billion or 16% excluding FX. Significantly exceeding the high end of our outlook. Net income was $1.9 billion, a 39% net income margin. Net income benefited from a $135 million reduction in our provision for income taxes as the prior year included a $213 million valuation allowance related to CMT Credit.

Adjusted EBITDA grew 22% year-over-year to $2.5 billion, representing an adjusted EBITDA margin of 52%, up from 50% in Q3 of last year. Margin expansion in Q3 was driven by strong revenue growth and cost discipline. -- including operating leverage from AI-driven efficiencies and customer support as well as the shifting of certain marketing expenses to Q4.

Let me take a moment on monetization and take rates. Our implied take rate was 17.6% in the quarter, down slightly year-over-year. That was primarily due to the expansion and strength of Reserve Now Pay Later, where guests book further in advance, increasing GBV before we recognize the associated revenue and to a lesser extent, higher incentives and credits related to new business initiatives. Partially offsetting these pressures to implied take rate, our guest travel insurance and our migration to single service fee. Guest travel insurance continues to build momentum. Revenue grew nearly 70% year-over-year with higher attach rates in every region.

Lastly, in October, we completed the final phase of our fee simplification initiative, whereby nearly all of our remaining hosts have now moved to a single 15.5% service fee. We've already seen that the single service set helps post price more competitively and creates greater transparency for our guests.

Next, our balance sheet and cash flow. We continue to benefit from our efficient and capitalized business model delivering $1.4 billion of free cash flow in Q3, up modestly year-over-year. We expect that free cash flow would have grown more in Q3 absent the impact of Reserve Now Pay Later bookings on working capital. RNPL defers guest payments from the time of booking closer to the date of stay impacting the seasonality of other needs. Free cash flow during Q3 was also impacted by a $75 million increase in purchases of property and equipment driven by our new office building in New York City. Over the trailing 12 months, we generated $4.9 billion of free cash flow, a 35% free cash flow margin.

Finally, during Q3, we repurchased $700 million of our common stock enabled by our strong balance sheet and significant cash flow generation, leaving $2.7 billion remaining under our existing authorization. And today, we're pleased to announce Board approval of a new $12 billion share repurchase program on top of this existing authorization. This reflects the continued confidence in our business performance, our ability to generate strong free cash flow as well as our commitment to returning capital to shareholders, which remains important to our capital allocation strategy.

Now let's turn to Q4 and full year 2026 outlook. We're encouraged by the meaningful acceleration we've delivered as well as the momentum we're seeing into year-end. Although we're closely monitoring ongoing developments in Europe, underlying global travel demand remains healthy and our faster pace of product innovation continues to drive results and our Q4 outlook reflects it. For Q4, we expect revenue of $3.22 billion to $3.28 billion, representing year-over-year growth of 16% to 18%. This include an approximate 2 percentage point FX tailwind after factoring in our hedging program.

We expect our implied take rate to be up slightly year-over-year. We expect GDV growth in the mid-teens, driven by at least low double digit -- low double-digit growth in Nights and Seats booked and a modest increase in ADR.

In Q4, we are not assuming a material persistent impact related to the 10 in Europe, specifically France or the complex in the Middle East, but are closely monitoring trends in both regions. On profitability, we expect adjusted EBITDA to increase year-over-year and adjusted EBITDA margin to be slightly down compared to Q4 2025, partially due to the timing of marketing spend that is earlier.

Now for the full year, once again, we're raising our revenue and adjusted EBITDA margin outlook despite facing more difficult comparisons as we head into Q4. We now expect year-over-year revenue growth in the high teens up from teens last quarter and our expectation at the beginning of the year of low double-digit growth. And we now expect full year adjusted EBITDA margin of approximately 36%, up from at least 35.5% last quarter.

During 2026, due to the expansion of Reserve Now Pay Later and its impact on the timing of working capital. We expect full year adjusted EBITDA margin to be modestly above our free cash flow margin. To close, Q3 demonstrates that the product and business changes we've made are translating into real financial outcomes, faster growth and expanding margins. We're encouraged by the progress we've made so far and the durability of our results.

Lastly, our product road map is robust and gives us a clear path to driving growth into the future. With that, I will open it up to Q&A.

Operator

[Operator Instructions] Your first question comes from the line of Richard Clarke from Bernstein.

分析师问答

Richard Clarke

I guess, firstly, on dynamic pricing. I guess in Q2, Brian, you said you thought dynamic pricing was a bigger opportunity than Reserve Now Pay Later. You've been out with that product for about a month now, just early learnings across the sign-up you're getting from host? And do you still stick with that? And just very quickly early, I think you said take rate up in Q4, why the reversal there? Is that because you're now lapping Reserve Now Pay Later? Or is there any other initiatives that are adding to the take rate in the fourth quarter?

Brian Chesky

Richard, so I'll take the first part. So yes, well, I think generally, pricing is probably one of the biggest levers we have, certainly in pricing. Just to give you a quick anecdote. A couple of months ago, we hired the Pine Rivers from Booking. He's our new Chief Booking Officer. And he was really instrumental in the early days of booking. And one of the things he said when he joined, he said, if you think Reserve Now Pay Later say, "Wait, you see we can do a pricing. It many Reserve Now Pay Later -- and it actually kind of makes sense. If you sell a product, customers care about 2 things: how good is the product and what is the cost? So it's really 50% of the buying equation and affordability is portion bearing because we have a lot of young travelers. -- it's much more economic stay-home, -- we're very much gen-based. We don't need data to share today about dynamic pricing, but the adoption is really great.

I think one of the things we've noticed, this is a really key point is that -- there is this win-win where dynamic pricing can generally lower prices across the marketplace. -- it still now host to make more money. So some nights, they make more money other than they say less. We were really able to find these cities where if you just charge like $10 less, you'll make enough more bookings that you'll actually make a lot more in the long run because your occupancy is much higher -- and that is like not an intuitive thing and it's good for the guest because it makes it a little more affordable.

And so it's a really huge lever. It's extremely early. This is something that like AI is really good at doing. And the thing about AI is like you need to tune it and get more data. So I think in the coming quarters, we should be able to have some more results. But again, I'll just kind of summarize by saying pricing is a massive lever and actually dynamic pricing is 1 of many things we're doing around pricing. .

Ellie Mertz

Great. And Richard, to answer here, take great question. about Q4, just been double in Q3 to explain the changing dynamic of or--are in Q3, we got an outsized impact of Reserve Now Pay Later on the gross booking value in quarter whereas obviously, the revenue from those bookings is delayed. And so in Q3, you see the impact is Reserve Now Pay Later increases denominated or bringing down the implied take rate at similar either versus in Q4, you should expect a modest increase in my take rate.

More broadly, I would say from quarter-to-quarter, there are some small timing movements that do impact take rate. But I want to call out as a single service fee as well as insurance should provide over a longer time period, a modest increase in terms of the overall rate.

Operator

Your next question comes from the line of Lloyd Walmsley from Mizuho.

Lloyd Walmsley

A couple if I can. First, just on the hotel effort. Can you kind of give us an update of how growth here is performing relative to the core, maybe or any other helpful quantification? And sort of where do you think this could be over the next few years as a tailwind?

And then second one, I appreciate all the color about agents in the shareholder letter very helpful. Curious, Brian, do you sort of see a path to your agent helping customers with travel sort of beyond the walls of Air DMD? Or do you think it's more limited to taking action and sort of managing services that you all provide yourself? Like how do you see that evolving over time? I know it's early, but curious to get your vision on that as well.

Brian Chesky

Yes. Okay. So hotels and agents beyond the wall even hotels are growing markedly faster than our core business. That's not another core business growing slow, just the hotel is already $900 billion market off of a very, very tiny base. I certainly don't want to put any like projections out there or forecast. But it's undoubtable that we will do multiple billions of dollars in revenue in hotels in the coming years. And we don't even need to like win this market to be a significant player and make billions of dollars.

And part of the reason why is we have a huge amount of traffic most people come to Airbnb are looking for a place to stay. If somebody doesn't find a place at they're probably going to another app or a website booking hotel. So -- we don't have to win. We just have to get the people in our store to buy some before they leave. And they know the -- we know they want something, they're in the store in the first place.

And the other thing is we're seeing -- this next thing was very surprising to me. every , when we started Airbnb, I was a little ideologically like not -- I was kind of against hotels. Let's be honest. My first half was forget hotels. And so we had a little animosity with the hotel industry, like in 2010, 2011, 2012, over the last 6, 7 years, really changed my tune approach. I'd like to think of this as growing up a little bit and being practical. And our guests told us, hey, we love hotels.

We bought hotels tonight before the pandemic. But we really decided to get serious about hotels a couple of years ago. And basically, one of the things we learned is that actually most people like staying in homes and hotels, it just depends on the type of trip. And that we don't have a campaign about why people stay at home, but if you're booking last minute for business travel, one night, hotels are really great. So we expect people to want to book hotels.

The thing I wasn't expecting was the reception from the hotel industry. And it was a complete 180 right Bank Hotel suddenly saw Airbnb a major ally. They really love the way they were merchandised. They love the fact that we were entirely new audience. We were disproportionately Gen Z. We are not just really affordable travel. We're actually really high-end travel because we have really big bills we are a much more global audience and they just like the factor is this for another channel.

So all that is to say, when we got really serious about hotels, it's not the only reason I hired Pine Rivers, but it was definitely a big thing. I told the Pine, this is a major opportunity for Airbnb. And it is going to be huge. And by the way, Pine Rivers ran the largest marketing organization ever at Booking.com. Now, I don't think performance marketing is a very important part of our core business, but I think with hotels. It's a commodity.

I think we're going to be able to do performance in really efficient because hotels are so high ABR Booking.com shows it's very profitable. But the best thing is we don't need to do a lot of it because we have so much organic traffic already coming in. So all of which to say it's going to generate many billions of dollars of revenue. I don't know what the time line is, but it's a huge opportunity. With regard to agents going outside the wall there yes, of course.

We are going to be -- that will be through MCPs. We're going to be integrating a number of MCPs with our trip agent in the first half next year. I won't give away too much more. We like to do these product launches. We don't want to build a general agent. Just to double click for 1 second. The problem -- if I just want to double click on a second, 1 problem with a general agent is you have to create like a general interface because the agent has to be able to do everything. So the general interface generally becomes almost entirely text based, right? Because we can have about travel medical information deliveries, you can't know what they're going to add.

As you get more specialized, you can develop a really custom interface that is perfectly bespoke for that type of use case. So with travel, working with the design these really custom components. If you want a car rental, the agent generates this little UI of a par and it's like beautifully in like a custom-generate interface. If you want to invite co-travels the agent helps you.

So there's going to be, I think, an incredible user interface experience that what I would describe as a kind of post chat-based interface. It's going to have and element to it. And so that's what we're going to be doing. It will be able to go outside the walls of me, but it's going to stay within travel and expect in the first half of next year.

Operator

Your next question comes from the line of Ken Gawrelski from Wells Fargo.

Kenneth Gawrelski

I appreciate it. Brian, I appreciate the commentary and the prepared remarks in the shareholder letter around the agent launch -- could you talk a little bit more -- and you just kind of elaborated that it actually could leave the walls of Airbnb -- can you talk about -- are you talking about actually direct -- maybe direct integrations with airlines and et cetera, to kind of complete the full trip.

Because I guess the question is for a vertical agent, if you don't have all supply, you don't have all attractions, you don't have all hotels. How do you think about the trade-offs for the customer between that horizontal general purpose agent, which may not be ideal for travel, but at least it sees everything. That's question one.

And two, the second one, please. if you don't mind, could you just -- and I apologize if I may have missed some of this. But could you talk a little bit just about the -- what you're seeing with RNPL as you continue to expand there. could you give us some confidence around cancellation rates and anything you knew you might be seeing there? .

Brian Chesky

All right. Yes, I'll do the AI agent, and then I'll hand over to Ellie for Reserve Now Pay Later. So the short answer is, yes, we are going to do integrations with a lot of providers. I don't want to give away too much again about our kind of product road map -- but I think it's safe to say we're going to do every part of travel. There's not 1 part of travel that we're not eventually going to do.

And every part of travel will have MCPs. We will have APIs -- and eventually, we do anticipate all bookings going through the app for every vertical and travel. I don't want to put a time on it. Just to give you an example, like say car rental. Right now, car rentals is done through a partnership called [ Troller ] and it's what we call link off. So an API but as a link off. The next phase is to the reservation booking directly on Airbnb. And so that's what we're doing with a number of these partnerships. So the kind of way we're going to do a lot of these partnerships is to get mass scale. We're kind of doing like link off. The link offs appear to feel like Airbnb.

So it's really in a -- like a web view that's designed in our customer interface, a with groceries -- this is with Legas storage, it's a car rental. And over time, we're going to develop a much more native booking integration. And obviously, the good of that is it's a little more seamless. You take a little bit greater cut of the revenue, but you also to handle customer service. So you have to be sure you're ready for that.

Now the great thing of it linkup is we can kind of decide and test what things we want to integrate when. But the end game is yes. The end game is eventually, we like all travel to be able to book natively through the app of Airbnb, but we're not going to put a time line on it. And then the other part of your question was general versus horizontal, right? Okay. So with regards to general versus horizontal and trade-offs, I mean, listen, there is going to be a reason for general agents. I'm not saying that there won't be general agents, but I want to just give my view.

I don't think I think a lot of people are looking at general agents like and they're trying to connect it to Google. Like there was 1 home page of the Internet. It was called Google and you went there and everything started at Google. This is my opinion. I do not think that's going to happen with chatbot. I don't think there will be 1 chatbot that's like Google in the front page, and it is the orchestra conductor for every other app. I'll tell you a couple of reasons why that I don't think that will happen.

The first problem is if that were to exist, I think it would have to exist to say operating system level. It exists at the Android operating system level at the MAC iOS operating system level. The operating system is the right substrate to control your apps. You get one app was the home page. You're really saying an app is going to control all the other apps.

By the way, you really think Apple is going to allow an app to control their entire computer. So we're considering all the potential issues for privacy security and that they can actually cut out of their own like user experience. I'm actually quite skeptical of that. So I do think that general SAT agents are going to be successful in popular. I just don't think they're going to be do everything.

And the burden of a general chat agent doing everything, they basically almost become an operating system. That is a lot to put into an app, and I just don't think that's going to happen. So I do think general Chase will be popular for single everyday things. I think as they get really, really complicated, specialization is going to win. -- now the changing can hand over to a specialized agent, but a lot of people are going to go direct to the specialized agent, just like they discover Airbnb on Google, and they come back direct. They don't just keep going through Google. That is generally my thought.

Now a general agent does have a bunch of data because you use the general agent more frequently, right? At the same time, we can connect MCPs of Gmail and Google Calendar. If you connected MCP of Google Calendar in Gmail or other MCPs, we can actually pull in what the users can send a ton of other data. So actually, there's many ways we can get as much data or nearly as much data or nearly as much data as a chat agent use every day. And this is just through ASIC MCP handoffs.

So I think basically, the way to think about it is -- in order for specialization to win, it must not be 20% better. It must be like 2, 3, 5x better than a general agent at its task because you do have the inertia and the switching costs. So our job isn't to be 20% better, it's to be multiples better. And that's the bar we'll maintain, too. We'll see if we can get there, but I'm pretty confident.

Ellie Mertz

Great. Ken, let me tell you just a little bit more about what we're seeing on Reserve Now Pay Later. Just as a reminder, Reserve Now Pay Later, we began to launch a year ago in the U.S. We expanded amongst our plus moderate inflation policy listings at the end of the year to retool. And then most recently, at the beginning of Q3, we expanded eligibility of Reserve Now Pay Later to limited and firm penciling policies. .

In doing so, it obviously grew materially the total eligibility of GBV on the platform for this product. As of Q3, it was over 40% of GDP was eligible to use -- in doing that expansion, we've seen a couple of things. First is a material lift to bookings. In particular, in Q3, we saw a full 3 points of net night lift associated with both the expansion and the PASeligibility offering. This was an incremental point or more relative to the impact we saw in Q2. candidly, that the expansion in the summer was even better than we had tested in terms of the reception for consumers. I

think it's probably pretty obvious, but this product make traveling more affordable, I need to pay upfront. And what we see is that people not only before they book earlier, so our lead times have expanded and they booked more expensive listings, which is one of the reasons that ADR is up. Specific to your question, cancellation rates in aggregate are up 200 to 300 basis points from a year ago. And this is -- if you think about the novel levels a year ago, our cancellation for order of magnitude about 16%, they're now about 19%, which we think is very healthy in the travel phase.

The final comment I will make, and I'm sure we'll get questions on it later, is effect of NPL in Q4. I just want to remind everyone that we had called out last quarter that Q3 would be the peak in terms of net impact of our IPO on our numbers. that's because we get the growth benefit in terms of the list of bookings before we see the full cancellations. And so the net impact to Nights booked will come down sequentially in Q4.

Operator

Your next question comes from the line of Justin Post from Bank of America.

Justin Post

Maybe one for Brian and a quick one for Ellie. Can you remind us what percent of your inventory is really proprietary for Airbnb? And how have you thought about using that as an asset, maybe with agent negotiations? How do you really take advantage of that. And then Ellie, just quick on the insurance. Can you give us anything on what the penetration is of bookings? Or how much is contributing to take rates?

Brian Chesky

Yes. Over 60% of our inventory, we believe, [indiscernible] Airbnb. And so that's the first part of the question. What's the other part of your question? It's the -- penetration of insurance products?

Ellie Mertz

Yes. Insurance products are growing quite nicely. It's a single-digit percent of bookings that have an attached that is up materially from a year ago. I think what excited on our insurance products is not only the growth, but also the opportunity to provide more insurance products. We're looking across a host of offerings that we could provide both guests and hosts and I should note that the margins are great on the attached insurance product.

Brian Chesky

Sorry, Justin, you said how does that factor into agent negotiations, exclusive inventory. Can you elaborate on that?

Justin Post

Yes. How are you thinking about that as an asset, either for your own internal agents or maybe in negotiations with external agents getting good financial terms? Or any strategic advantages there? Yes. .

Brian Chesky

I'm not sure yet. I mean I think it's really -- I think as far as speculating what the relationship is between unique inventory and like potentially leverage, negotiating with like other partners. I think it's really early, I don't know. But let me just tell you my philosophy on unique inventory. I don't pay a lot of attention looking at what percent of our inventory is unique anymore. It's helpful, the more unique to better. There's no question. .

The most important thing is as we go into categories, we are -- our philosophy is it's okay to have commodities, like hotels. It's okay to have some cross list at home. -- as long as there are a lot of unique things that you can't get anywhere else. And I think the percentage is not as important. It's the absolute amount, right? So like if -- imagine there are 2 stores, they all have the same products except 1 store has a bunch of other stuff that other store doesn't have.

The percentage of unique stuff in the store doesn't matter just how many things, the items are unique in that store, the other store didn't have. And so as a theoretical mental model, Imagine there's 2 stores and 1 they have almost all the same inventory, but 1 has just some more unique stuff. If you have 10% more up, you might get more than 10% more traffic because suddenly, it's a major difference. So we don't really look at the percent mix. I mean it's good to know.

Of course, we track it, but we are growing really, really quickly. by process of inventory and getting that really competitive nice. And actually, 1 of the things noted where now pay later, is this helped us become way more competitive like getting those nights. And then the hostile fee has been really competitive with us keeping more competitive prices. And so why I say this is just to finish off this point.

With services, a bunch of our inventory is commodity inventory, like car rents. You can get those cars elsewhere. But then we have a bunch of inventory you can't find anywhere else, like misuse, chefs and photographers. As experiences, we're now offering ticket. You can get about those tickets on their website, but we also have exclusive experiences. So kind of category by category, we'll get a bunch of the commodities. We always want to make sure that we have a unique stuff. And we have unique plus commodities, then we become the one-stop shop. You don't need to go or in or out.

Operator

Your next question comes from the line of Benjamin Black from Deutsche Bank.

Benjamin Black

Great. So Brian, it seems that you're increasingly adding social and travel discovery elements through connections, through Travelmat through neighborhood pages even personalized recommendations. So how do these features change the role you guys play before guests is actually those destination or before they're ready to book. And could greater discovery eventually lead discovery element eventually lead to sort of a more natural opportunity for sponsored listings? And then secondly, as you add hotels, you expand a number of experiences, you add services, how does that change your thinking around loyalty more bodily .

Brian Chesky

Two good questions. So social and travel discovery, let me just like before I answer the question, let me back up. So I remember like 22 years ago when Facebook first came out, it came out in 2004. I couldn't use it right away because I just missed it. I just graduated from college right before Facebook launched in my college, I had to wait 2 years before they open up to everyone. So I remember Facebook would used to be a social network. And with basically your friends and you can keep in touch with people and you can message them and see who they were doing. And that was around 2004 to 2006, 2007, 2008, let's call that the golden age of social networking.

And at some point, social networking morphed into social media. And our friends became our followers, and we stopped connecting and we started performing. And a very bizarre phenomenon happened where the social network in a sense was invented in kind of for a large part of the world, uninvented, like I no longer have a social graph. I no longer know whose birthdays, my friends are unless they're close friends. I no longer know when somebody -- like something happens in someone's life, I no longer know how to keep a touch with people, except I got to message them and with that for my messaging platform.

So there really aren't social grafts anymore. I mean, LinkedIn is kind of a resamegraph, but really, there's no social grafts anymore. And we thought what if we brought back the social graph, but instead of being an online social graph, -- it was a social graph in the real world of the deep you travel with and the people who meet along the way of your travels. And we thought this could be really interesting.

So we started building these social travel features this year. We launched them in May. We launched a travel map -- and we started realizing this might be the beginning of something massive. And we are going to be launching a lot more social features and social connected features next year. I think this is maybe one of the single most underestimated parts of our business.

I mean, it might be because we don't talk about them much, but that's okay. I don't want to give away too much of the road map, but I can just say a couple of things. Number one, it's a great proxy for trust. If I were a host, and I were to get a reservation for somebody, I don't know them, but I can see how they're connected to me, how many degrees away, that would be a really good proxy for trust. And as you build this social graph, you increase that as you can trust you potentially go into a lot of the businesses. I'll let your imagination run wild what's possible.

You can also see where your friends have traveled and stayed. So you can see like my friends at Paris, now I can message them and ask what's going to do in Paris. I can eventually -- we can eventually bring that into search results. Right now, that's on the profile have, but we eventually are going to bring that into searches. So when you see the map, you'll be able to see where your friends have travel in search results in map. You'll be able to see listening to friends of stage. You'll be able to see who's been traveling there recently and where they're going. So there's going to be a lot of these features.

These features are going to, a, make things more sticky, b, it's going to make group travel much more collaborative. -- see, it's going to make it much easier to get great recommendations that's going to increase trust in the platform. And the other thing I'll just say the last thing is, I've been doing -- I occasionally do these roundtables and with customers.

And I recently did a Gen Z roundtable, basically kids that are like college age. And they were kind of like a little younger than the age I was when I started Airbnb. And I heard 2 things from Gen Z. And again, I'd just Gen Z is like the fastest-growing segment, and it's the prime advertising segment. They basically 2 things. One, they want to save money. Young people always want to save money. -- but in age of rising cost of living with a lot of college debt with tuition going up, and the latter being pulled up a little bit with AI, a lot of entry-level jobs are harder to come by because AI can do them more than ever, they're looking for affordability.

But even more so, we're hearing a lot of college and recent granting they want to meet people. And a lot of people, they meet people at work in school, search, Boeing alleys. Well, a lot of these things aren't around as much anymore. And a lot of kids if they don't have a job, they're not going to be some of the work. So we're starting to see that a lot of these social features are going to be especially popular with Gen Z.

But surprisingly, we're also seeing like on the other barbell empty nester, Kids are out of the house. They want to travel. They want to meet people. So there's a lot of opportunities here with the social features. With regards to sponsor listing, the relationship between them, I don't know. I think the thing I'll say about sponsor listings in particular is we are developing our AI agent, and we wanted -- we are developing sponsor listings, but they'll be developed in the context of the AI agent. So we first need to launch the agent.

We need to refine it, and then we figure out how ads work in that new world. By the way, everyone's figuring is out at stinging this out. So I don't think it's a clear -- it's not clear what the implementation is -- but I think if anyone can figure out, I think our design team side. With regards to a loyalty program, I mean, I don't -- again, I want to not break any news, but we are absolutely looking at a loyalty program.

We've been looking at it for a long time. I think I've said for a long time, the best loyalty program is plus customers loving your product. Disney and Apple for years. lap doesn't have a loyalty program. People love their products. But at the same time, we're like the only travel major traveler in the world that doesn't have one.

And I think we have been at a disadvantage because we don't have a loyalty program. A lot of people want to use the say, "I want to use you, but you don't have a program. So I almost feel like I can't use you." -- so we are absolutely looking at a loyalty program. I don't have anything announced, but I can say that as we add more verticals -- we just have a lot more to work with. So adding verticals make loyalty program more compelling.

Operator

Your next question comes from the line of Jed Kelly from Oppenheimer.

Jed Kelly

Just looking at the guidance, it looks like your EBITDA margins are going to be in that 36% range, so above 35%. But if I look at your growth, you're growing 12%. I think, Ryan, knowing you, you like faster growth. So can you just talk about how we should think about threading the needle between double-digit room night growth and maintaining margins above 35%, 36%?

Ellie Mertz

Jenny, are you talking about Q4 or '27, what's the reference?.

Jed Kelly

Just balancing profit versus growth given that you're seeing accelerating growth and what your preference is?

Ellie Mertz

I mean I think Brian and I have been very clear over the last 2 years that we have very been prioritizing growth focused on driving growth in this business. And I think the results in the last couple of quarters show that we have been able to deliver that growth. behind that, we continue to invest in growth. But at the same time, in the last couple of years, what you've seen is that we've been capable of reinvesting upside in order to continue the momentum. .

We're obviously in the '27 planning right now. And the prioritization around continuing this momentum continues to be the top priority -- the good news is that this business is extremely profitable, has incredible cash flow dynamics that give us a lot of room to invest within a very strong profitability envelope.

Operator

Your next question comes from the line of Matthew Cost from Morgan Stanley.

Matthew Cost

I guess, firstly, on the hotels and the people that are using the hotel product, -- are they generally new to Airbnb and bringing in incremental bookings? Are they a high-end or low-end consumer? How are they retaining? I guess, tell us a little bit more about the behavior of those people that are booking hotels -- and then secondly, Brian, I was wondering if you could just respond to some of the data that's out there about early adoption of these agents that's showing agents like news and instinct that's showing that a lot of commercial activity happening there is travel related. Is that something you think strengthen the mix over time? Or how does that gel with your overall view of how people will use these horizontal tools?

Brian Chesky

Maybe I'll start with the second one and just to mind then Ellie you can go to the first one. So we're not seeing a lot of traffic in even fewer bookings from us and instinct, yet, we do have nearly 0.5 million bookings a day and a very nonmaterial number of bookings happen I'm using instinct. So we're talking very low numbers. So we're not seeing a lot of traffic today and it's not really increased over the last 4 weeks.

So it kind of peaked in early October. It's kind of been stable in the last 4 weeks. So we're kind of watching it. My general philosophy on agents are -- I'm generally excited about them. I don't see them as a threat. I see them kind of like Google as an opportunity for more people to discover Airbnb. So I think they'll be great handoff to us. There will be people that will want to book a one-off flight or something directly on an agent.

So I'm not saying that will never happen. But travel is -- maybe this is like just a framework to think about. There's item-based retail some project-based retailers. So just as an analogy, -- you go to Walgreens. You don't go -- you don't like plan to go to Walgreens. You need to get something to go to Walgreens. You go to IKEA or Home Depot, that's a project-based retailer, right? You have a whole project. You don't go to those places to get on it.

Airbnb we're really long term, we want to be in the midst of the entire trip. That's entire project. It involves where stay, what to do, how to get there all in place. It's highly logistical. Our guests typically travel in groups of 3 or more. 90% of people message the host. The host wants to know who's in their home, so they want to have a connection. They don't want to offload to an agent. If I'm hosting somebody, I want to talk to the person, not the agent on behalf of the person -- it's very visual.

People typically look at multiple -- dozen listings or more. They look at 25 photos per listing often. So it's very visual. It's very collaborative. It's a very complicated experience. And the other thing I mentioned this in my opening remarks, is there's a lot of studies that show people enjoy anticipating travel more than actually traveling. Here's a fun fact.

Most of our traffic happens during 9 to 5 hours. These people are supposed to be working. And they're not. We have evidence. People are actually secretly using Airbnb. Maybe they're trying to skip their job. But the idea is that it's highly aspirational. So I don't think a lot of people want to offload all that to an agent. I think they want to offer the things they don't enjoy not the things they do enjoy. So it's small.

We're watching it, but we are definitely open to working with these agents. But I also just think they'll be really helpful handoffs to us. But I think that to do -- I'm not saying agents can't do travel, I don't get to focus on travel, and that's what we're going to do. We're going to focus on travel. And any want to focus on travel can probably solve it, but a general agent trying to solve travel and solve 50 other things. That's really hard.

Ellie Mertz

To have some questions, I'll be quick, and then we're going on time. In terms of the mix of nevers versus past bookers, it's actually very consistent with our overall platform. We're seeing great demand on both segments. And -- and just to underscore why like that is in with our strategy. We're eager to get more travel share from our past guests and hotel spoken up new use cases that even our most loyal guests may be choosing hotels of the platform. .

Similarly on the first-time booker front hotels part of the trend. -- to onboard those guests that haven't yet tried for a home by operating them something they are familiar with, and we're seeing that as well. So great hotel demand on both sides of the marketplace.

Operator

And we have reached the end of our question-and-answer period. I will now turn the call back over to Brian Chesky for some final closing remarks.

Brian Chesky

All right. Well, thank you, everyone, for joining. I just want to end where I started. Airbnb is back in growth mode and it's here to stay. This is a result of years of hard work to make every better. And as we build by to become an AI native company, I actually think the opportunity ahead of us is actually even bigger. I want to end most importantly, by thanking our incredible team. I get to I get -- we get to take the credit when we have good calls, but it's only because of the great work of the people.

And I think one of the superpowers of Airbnb, maybe the Super Power, this is old cliche, it's all about the people. But it really is, especially in Silicon Valley and tech because you're only as good as your most recent innovations, and that's only as good as our team. Our team tirelessly. -- they've been really, really burning the many oil a lot of times this year.

And so I'm asked why are we growing so fast. I often say things like execution. Well, execution only happens from people teams. And I think that what these results really are beyond AI, beyond all these products, is this is a team that's really starting to get into flow. They're doing great. So thank you all, team Airbnb. And thank you all for dialing in today. We are just getting started back in growth mode. Thank you.

Operator

This concludes today's conference call. Thank you for your participation. You may now disconnect.