Booking Holdings (BKNG.US) Q3 conference call: CFO says the Q4 deceleration has been factored into the impact of the Middle East conflict, and the average number of daily AI search visitors increased tenfold from month to month

The Zhitong Finance App learned that in the early morning of November 4, Beijing time (after market on November 3 EST), Booking Holdings (BKNG.US), the world's largest online travel company, announced the results for the third quarter of 2026 and held a results conference call after the market on the same day.

Total bookings for the third quarter were US$54.5 billion, up 10% year on year (fixed exchange rate caliber increased by about 11%), higher than market expectations of US$52.69 billion; overnight volume was 345 million, up 7% year on year, better than the forecast of 339 million. GAAP revenue was US$8.89 billion, down 1% year on year, below market expectations of US$9.54 billion; however, GAAP revenue for the quarter included an FTC-related revenue reduction of US$900 million. After excluding this matter, the adjusted revenue disclosed by the company was US$9.79 billion, up 9% year over year (fixed exchange rate increased by about 10%), higher than market expectations. GAAP net profit was $2.7 billion, down 1% year on year; GAAP EPS was $3.61, up 7% year over year. Adjusted net profit of $3.5 billion, up 8%; adjusted EPS was $4.66, up 17% year over year, higher than the forecast of $4.49. The adjusted EBITDA was about US$4.6 billion, up 9%, higher than the forecast of US$4.49 billion; the adjusted EBITDA margin was 47.0%, the same as the same period last year.

In terms of stock prices, the stock closed up 1.25% to $155.67 during the regular trading session, and the stock price declined after the market. Since this year, the stock has fallen by about 27% cumulatively.

Market Focus and Management Response

There are two superficial “contrasts” in this financial report: one is that GAAP revenue falls short of expectations, and the other is the overall deceleration of the Q4 guidelines. The decline in GAAP revenue stemmed from a one-off event — the company cut revenue of $900 million and accrued losses of $100 million in a US Federal Trade Commission (FTC) related matter, both of which have excluded adjusted metrics. Notably, management did not discuss or respond to FTC matters during the entire call, and no analysts mentioned it during the Q&A.

On the call, management responded positively to the Q4 deceleration issue. CFO Steenbergen gave Q4 guidelines in his press release: overnight volume increased 1% to 3% year over year (the strictest year-on-year basis for 2026), total bookings increased 2% to 4% (fixed exchange rate 4% — 6%), adjusted revenue increased 0% to 2% (fixed exchange rate 2% — 4%), adjusted EBITDA remained roughly flat year over year (fixed exchange rate increased by about 2%). “We currently expect nighttime volume growth of 1% to 3% in the fourth quarter, which reflects the tightest year-over-year base in 2026, as well as uncertainty about the impact of recent events on travel demand,” he explained. Our guidelines reflect these recent disturbances and assume that both the direct and indirect effects of the Middle East conflict on travel will continue until the end of the year.” He said that global travel demand had cooled down at the end of the third quarter, with international trends weakening and domestic resilience; demand in October was affected by the suspension of round-trip travel to the Middle East, the conflict indirectly boosted ticket prices, and local turmoil in some European cities.

The questions in the Q&A session all focused on AI topics: from quantifying the actual effectiveness of AI, cooperation with platforms such as Meta, OpenAI, Google, Anthropic, and Amazon, testing two native AI brands, LoLA and Vayo, to the impact of EU DMA regulations on the popularity of individual AI agents.

AI: Management gives four categories of early performance indicators

In response to a question from Douglas Anmuth of J.P. Morgan Chase Research about the actual effectiveness of AI, Steenbergen gave four categories of “positive early indicators”:

The first is increased engagement: “Booking.com's average daily AI search visitors surged tenfold in the third quarter.” The second is an increase in conversion rate: “Priceline and Agoda's AI assistant users have a conversion rate of about 2 times that of non-users.” Third, the reservation value is higher: “The average reservation amount for Agoda AI Assistant users is about 20% higher than that of non-users.” Fourth, incremental: OpenTable's native integration with OpenAI “has become one of the fastest growing channels for the brand, generating 190,000 seat covers (seat covers) in September”, and OpenTable's interaction with Muse and other horizontal agents “brought a large number of new customers” to our platform.

Regarding LLM and AI agent drainage, he said, “Overall, this type of traffic is still very small... still significantly less than 1% (overnight ratio)”. Although there has been an acceleration in recent weeks, “no channel is growing significantly faster than other channels.”

AI Strategy: Two AI Native Brands “Test Field” Cooperate with All Platforms

CEO Glenn Fogel said in a press release: “We believe this is the most transformative era for the industry since the Industrial Revolution.” The company launched Lola in September — a conversational experience and membership model to help users obtain unique travel, dining, and event inventory outside of the company system; a few weeks later, it also launched Vayo — an active AI personal travel agent that can aggregate existing reservations in travelers' emails, continuously monitor prices, and maintain interaction after booking. He said these platforms are “testing grounds” (testing grounds), and their ability to succeed will be promoted to major brands such as Booking.com, Agoda, and Priceline.

In terms of cooperation, the company “continues to cooperate with leading technology and AI companies such as Meta, OpenAI, Google, Anthropic, and Amazon”; Agoda launched Muse in Asia as a starting partner, directly connecting to Agoda's new AI assistant (already covering 21 markets); and OpenTable became the first catering partner to connect to personal agents. When asked by Evercore ISI's Mark Mahaney why Agoda was the first to launch Muse, Fogel responded: “You shouldn't think Agoda's launch in Singapore is more important than our eventual partnership with Muse or anyone else in the US. Please do not misread it. We have a great relationship with all participants, and we know there are more partners on the way.”

Regarding Citigroup Research's Ronald Josey's question about the supply value of Booking, Steenbergen said that horizontal agents are willing to cooperate with aggregators because “connecting 4.7 million listings directly and obtaining real-time price and property status data is too expensive for them”, and linking with the Booking brand can bring “instant credibility”; Fogel added that if these agents move towards some kind of monetization model in the future, “our situation will be even better”.

Business segment: Faster alternative accommodation, strong air tickets and attraction tickets

Alternative accommodations accounted for about 37% of Booking.com's overnight stays, up about 1 percentage point year on year; overnight stays increased 7%, significantly faster than 4% in the second quarter, and accelerated across the board in major regions; housing volume growth accelerated to 9%, which the CFO called “an encouraging sign for future growth in this category.” Ticket volume increased 19% year over year, a sharp acceleration from the second quarter. “Despite high ticket prices and reduced capacity on some routes, it still outperformed the market”; attraction tickets increased by 24%. Connected Trip's transaction growth rate accelerated to nearly 20%, accounting for a low double-digit ratio of Booking.com's total transactions, and a growth rate of 3 times that of the company's overall transactions. The merchant model accounts for about 75% of total bookings, an increase of about 3 percentage points over the previous year.

In terms of prices, the fixed exchange rate ADR (average daily housing price) increased by about 3% year over year, mainly driven by Europe, and the US and Asia also contributed. In terms of channel structure, B2C direct sales have stabilized in the mid-60% range over the past four quarters; the proportion of mobile apps at night was 50% higher, higher than the mid-50% range in the same period last year; Genius Level 2 and 3 members accounted for more than 30% of Booking.com's active customers, contributing 50% to overnight traffic in the 50% range.

Cost and efficiency: AI is already cutting costs, and transformation plans are progressing on track

Total operating expenses decreased 1% year over year; marketing expenses increased 11% year over year, slightly faster than total reservation amount, mainly due to the faster increase in B2B affiliate marketing expenses; after excluding B2B affiliate business, B2C business achieved marketing leverage this quarter. After the adjustment, sales and other expenses account for 2.0% of the total reservation amount, and the increase in third-party customer service efficiency enabled by AI “is enough to offset the increase in payment costs brought about by the increase in the share of the merchant model”. Adjusted fixed operating expenses increased by 6%, with a slight decrease in personnel costs.

In terms of the transformation plan, the company “is still expected to achieve total annualized savings of US$650 million in 2027”, generating about US$23 million in transformation costs in the third quarter (most have excluded adjusted results). Examples of AI cost reduction cited by management: Agoda's mixed cost per million AI tokens fell by more than 25% between June and September; customer service costs declined and satisfaction remained high as reservation volume increased; engineering efficiency improved, and the cost of a single merge request continued to drop by double digits. Steenbergen added that the company's workforce growth over the past two years was “extremely limited” and was mainly concentrated in competency centers in low-cost locations such as India.

Regarding Benjamin Black of Deutsche Bank Research Division's question about whether the “8% revenue/15% EPS” long-term growth algorithm will be broken through by AI efficiency, Steenbergen responded that “it is still too early to judge how AI will change the long-term algorithm”, but emphasized that AI “has moved from experiments to measurable operating value across the company”.

Return on capital: quarterly buybacks of $2.1 billion, and $10.5 billion has been returned throughout the year

Free cash flow of US$1.1 billion for the third quarter was affected by a seasonal decline in US$3.6 billion in deferred merchant bookings; cash and investment balance at the end of the period was US$16.2 billion (US$17.7 billion at the end of the second quarter). The current quarter was returned to shareholders of US$2.4 billion, of which US$2.1 billion was repurchased (excluding consumption tax), and the remaining repurchase authorization as of September 30 was US$12.4 billion. The cumulative return of capital in the first nine months of 2026 was 10.5 billion US dollars, of which 9.5 billion US dollars were repurchased. The board of directors announced a quarterly cash dividend of $0.42 per share (25 to 1 after stock split) to be paid on December 31.

Q4 Guidelines and 2027 Outlook

The Q4 guidelines have been described above. The exchange rate is assumed to be EUR/USD 1.12, and the exchange rate is expected to slow down the Q4 adjusted revenue growth rate by about 2 percentage points. Maintaining full-year guidelines: high single-digit growth in total bookings, adjusted revenue, adjusted EBITDA, and low to medium double-digit growth in adjusted EPS.

Asked by Justin Post of the Bank of America Research Division about investment in 2027, Steenbergen said that Q4 will increase “disciplined long-term investment” in non-residential verticality (air tickets, attractions, restaurants), AI capabilities, and technology infrastructure, partly offset by marketing leverage; “It's too early to comment on 2027, and we'll talk about it again when the fourth quarter results are released in February next year”; “But on-the-books (on-the-books) data for the first half of 2027 looks healthy, which gives us confidence in the outlook for next year.” . In response to Morgan Stanley's Matthew Cost's question about whether AI agents are generating incremental travel demand, Fogel said that fragmentation of the discovery process “may allow us to acquire customers at a lower cost”, and that Steenbergen added that channel diversification “ultimately benefits us economically”.

In response to Redburn Atlantic's Alex Bruno's question about the impact of the EU DMA, Fogel said: “Currently, we haven't seen any significant advantage from this change... I haven't seen any changes in numbers.” At the same time, he warned that regulatory frameworks such as DMA, EU AI Act, and DSA may cause “the spread of horizontal personal agents in some regions far slower than expected.” Regarding HSBC Research Meredith Prichard Jensen's question about loyalty networking, Fogel confirmed that the company is exploring loyalty program connections with partners within and outside the hotel and travel ecosystem to expand the value of the Genius network.

Full Meeting Minutes

Company executives attending the conference: Glenn Fogel (CEO), Ewout Steenbergen (Executive Vice President and Chief Financial Officer)

Questioning analysts (in order of questions): Douglas Anmuth (J.P. Morgan Chase Research), Mark Mahaney (Evercore ISI), Matthew Cost (Morgan Stanley Research), Alex Bruno (Redburn Atlantic), Justin Post (Bank of America Research), Ronald Josey (Citibank Research), Benjamin Black (Deutsche Bank Research), Meredith Prichard Jensen (HSBC Research)

The host opened

The moderator read out the Safe Harbor Statement (based on the Private Securities Litigation Reform Act of 1995), prompting forward-looking statements of risk, introducing company spokespersons Glenn Fogel and Ewout Steenbergen, and asking analysts to limit one question at a time.

Glenn Fogel (CEO) Preliminary Statement

Fogel said that the company's various teams have once again delivered strong operating performance. All key financial indicators have exceeded the upper limit of the guideline. The performance reflects the resilience of travel demand and the team's efforts to strengthen products and services. Although short-term travel trends may fluctuate, the company is still focused on making travel more convenient, connected, and personalized.

On the AI side, he said, “We believe this is the most transformative era for the industry since the birth of the Industrial Revolution,” and AI is pioneering new ways to discover, plan, book and manage travel. The company actively tests AI throughout the traveler journey, learning from actual use and expanding the most valuable capabilities. Booking.com is expanding AI tools to help travelers explore destinations in more detail, and is promoting agency-based capabilities so that customers can cancel reservations or get answers to itinerary questions without contacting customer service. Early tests showed encouraging customer satisfaction and reduced service team upgrade processing capacity. Agoda's new AI assistant has been launched in 21 markets to help travelers go straight from initial ideas to bookable itineraries; Priceline continues to develop its AI travel assistant, Penny, and incorporate more abilities into the main experience.

In September, the company launched Lola to help users obtain unique travel, dining, and event inventory outside of the company system with a conversational experience and membership model; a few weeks later, it launched Vayo, an active AI private travel agent that can summarize existing reservations in traveler email history to generate more personalized offers and suggestions to fill itinerary gaps, and continuously monitor prices, maintain interaction, prompt changes, and assist in adjusting plans after booking. These platforms are test beds that can be quickly iterated and experience applied to the entire Booking Holdings system.

In terms of cooperation, the company continues to cooperate with leading technology and AI companies such as Meta, OpenAI, Google, Anthropic, and Amazon. OpenTable's native integration with OpenAI has become one of the brand's fastest-growing channels; OpenTable is also connecting personal agents as a catering partner for the first time. Agoda has launched Muse as its first partner in Asia, connecting Muse directly to Agoda's new AI assistant.

The US and Asia are two key growth markets. On the US side, the growth rate of intercompany nights has continued to outperform the market over the past three years, direct sales interactions have deepened, and more travelers have booked trips on the platform. There are many Asian markets, and each requires localized execution. Booking.com's global coverage and Agoda's local expertise put the company in an advantageous position. On the supplier side, independent hotels and alternative lodging partners contribute the vast majority of nights on Booking.com — the top ten global hotel chains together only account for a low double digit ratio. The company provides partners with demand data and tools, as well as products such as Genius to help them enhance their competitiveness. Connected Trip deals (multiple vertical categories booked for the same itinerary) are growing at 3 times the company's overall rate, and such travelers return more frequently; Genius 2 and 3 members account for more than 30% of active customers, contribute 50% more night time in the range, and have earlier plans, more stable return visits, and a higher direct sales reservation rate.

Ewout Steenbergen (Chief Financial Officer) Preliminary Statement

According to Steenbergen, overnight volume, total bookings, adjusted revenue, and adjusted EBITDA all exceeded the third-quarter guideline limits, “achieving profitable growth in a complex macro and travel environment while continuing to invest in long-term priorities.”

Q3 results: Overnight traffic increased 7%, up from 5% in the second quarter. Driven mainly by improved international travel sources from Europe and Asia, the domestic overnight growth rate stabilized at a high level. Total bookings increased by 10% (fixed exchange rate around 11%). The fixed exchange rate ADR rose by around 3%, mainly driven by Europe. Adjusted revenue increased 9% (fixed exchange rate of about 10%), accounting for a year-on-year decrease of about 15 basis points in total bookings. The adjusted EBITDA was about 4.6 billion US dollars, an increase of 9%; the adjusted EPS was 4.66 US dollars, an increase of 17%, which was faster than the EBITDA growth rate, and was helped by a 7.5% reduction in the average number of shares.

Structural indicators: B2C direct sales have stabilized in the middle of 60% in the past four quarters; the proportion of overnight visits from websites and apps has increased year-on-year. The proportion of mobile apps at night was 50% higher (50% in the same period last year). The percentage of nights among Genius Level 2 and 3 members is 50% higher. Alternative accommodations accounted for about 37% of Booking.com overnight stays, an increase of about 1 percentage point over the previous year. Overnight growth increased 7% (4% in the second quarter), and all major regions accelerated across the board; housing growth accelerated to 9%. Ticket volume increased by 19%, a strong acceleration from the second quarter. Despite high ticket prices and reduced capacity on some routes, it is still ahead of the market; attraction tickets increased by 24%. Connected Trip's transaction growth accelerated to nearly 20%, accounting for a low double-digit share of Booking.com's total transactions. Total merchant bookings account for about 75% of total bookings, an increase of about 3 percentage points over the previous year; the merchant payment platform is a key support for Connected Trip's vision, providing partners with more than 100 payment methods, the ability to accept payments in more than 50 currencies, and contributing to incremental marginal profits.

Expense side: Marketing expenses increased 11%, slightly faster than total bookings, mainly driven by a faster increase in B2B affiliate marketing expenses; excluding B2B affiliate business, B2C business achieved marketing leverage this quarter. After the adjustment, sales and other expenses account for 2.0% of the total reservation amount, leveraged. The increase in third-party customer service efficiency brought about by AI enabling tools is enough to offset the increase in payment fees brought about by the increase in the share of merchants. Fixed operating expenses increased by 6% after the adjustment, reflecting cost control actions (including a slight reduction in personnel costs). The transformation plan is still on track, and the total annualized savings of US$650 million will be achieved in 2027; the transformation cost for the third quarter was about US$23 million, and most of the adjusted results have already been eliminated.

Q4 guidelines: overnight volume increased by 1% to 3%, reflecting the toughest year-on-year basis in 2026 and uncertainty about the impact of recent events on demand; total bookings increased by 2% — 4% (fixed exchange rate 4% — 6%); adjusted revenue growth of 0% — 2% (fixed exchange rate 2% — 4%), slower than total bookings due to the share of ticket business and differences in the timing of booking and travel, partly offset by increased payment revenue; Adjusted EBITDA remained roughly flat year over year (fixed exchange rate increased by about 2%), which will be affected by incremental investment in strategic priorities, including long-term ticket expansion, catering Non-residential vertical capabilities, promotion of AI capabilities, and strengthening the underlying technical infrastructure and platforms are partly offset by marketing leverage. In terms of the travel environment, global demand cooled down at the end of the third quarter, driven by international travel; domestic travel remained resilient, and travelers preferred shorter trips and closer destinations. In October, the suspension of travel to and from the Middle East, indirect effects of the conflict (such as high ticket prices), and local turmoil in some European cities affected demand; the guidelines have taken these disruptions into account and assume that the direct and indirect effects of the Middle East conflict will continue until the end of the year. Guidelines for maintaining total bookings, adjusted revenue, high single-digit growth in adjusted EBITDA, and low to medium double-digit growth in adjusted EPS throughout the year (reporting scale).

Cash and returns: Cash and investment balance of US$16.2 billion at the end of the third quarter (US$17.7 billion at the end of the second quarter); free cash flow of US$1.1 billion, affected by the seasonal decline in US$3.6 billion in deferred merchant bookings; return of US$2.4 billion to shareholders, including US$2.1 billion in repurchases; cumulative return of US$10.5 billion over the first nine months of 2026, including US$9.5 billion in repurchases.

Q&A session

Douglas Anmuth (J.P. Morgan Chase Research Department): The actual progress of AI in various brands - in addition to reducing customer service costs, can revenue side benefits such as conversion rate and frequency be quantified? Also, in the AI referral traffic that has accelerated in recent weeks, how are new agents such as Muse and Grokbot different from previous LLMs?

Steenbergen: Four types of positive early indicators - participation: Booking.com's average daily AI search visitors surged 10-fold in the third quarter; conversion rate: Priceline and Agoda's AI assistant user conversion rate was about 2 times that of non-users; reservation value: Agoda AI assistant users had an average booking amount of about 20%; incremental: native integration of OpenTable and OpenAI generated 190,000 subscribers in September, brought about by the interaction of horizontal agents such as OpenTable and Muse Tons of new customers. Referral comparison between LLM and new horizontal agents: “Overall traffic is still very small. None are growing significantly faster, and are still significantly below 1%.”

Mark Mahaney (Evercore ISI): Why was Agoda the first to launch Muse in Asia instead of other global brands; and B2B marketing deleveraging has continued for several quarters, and how is the B2B business progressing?

Fogel: The company is always cooperating with everyone, and various collaborations are being implemented quickly and slowly; you should not think that Agoda is more important than cooperation with Muse or others in the US because it was first launched in Singapore; don't misunderstand it; the relationship with all participants is very good, and there are more partners in the future.

Steenbergen: The B2B business is receiving more internal attention; the merged B2B business will be launched on January 1, 2027, and it is likely that the data will be disclosed separately at that time. The reason for the increase in B2B marketing expenses was the increase in commissions settled on revenue share (rev share) after the business grew, and “behind this is health — because rising commissions mean that the business is very healthy from an economic point of view”.

Matthew Cost (Morgan Stanley Research): Commercial activity on early AI agents was clearly biased towards travel — was this just a convenience for existing bookers for Booking, or did it actually generate incremental demand by reducing friction in complex purchasing processes?

Fogel: I do see the value mentioned. Many people use personal agents or AI for travel discovery, which is fun and easy to use, and ultimately leads to a large number of bookings; but as Ewout said, the proportion of referral traffic is still very small. In the long run, it may become another important customer acquisition channel, reducing friction and increasing conversion. The company's purpose of launching products such as Vayo is to make users think of coming to us first; to use AI to complete connected trips, provide more value to consumers, and let travelers “come to us first”. It was discovered that fragmented portals “may be a good thing, allowing us to acquire customers at a lower cost”.

Steenbergen: Strategically, the importance of new channels lies in the diversification of third-party channels, which ultimately brings economic advantages. Competition among AI platforms is fierce, and the business model has not yet been defined. “They prefer to connect to aggregators, which is beneficial to us”; channel diversification will eventually increase the company's positive area.

Alex Bruno (Redburn Atlantic): The composition of B2C marketing leverage; and how will the impact of mainland European DMA on Google search results display affect company data?

Steenbergen: The management method of B2C leverage is the same as before: one is to continuously increase the ROI of paid channels, and the other is that direct sales traffic is growing at the same time. The combination of the two brought B2C marketing leverage in the third quarter; it is also expected that marketing leverage will continue to be achieved in the fourth quarter.

Fogel: Currently, I haven't seen the DMA change bring a clear advantage. It's still early, and there are no changes in numbers; “Some people say this will benefit OTA a lot; I hope so, but I haven't seen it yet”. Using this, he reminded an issue that few people have discussed: the global regulatory framework — DMA, EU AI Act, DSA, relevant UK regulations, etc. — will determine what kind of access to personal information can be obtained by horizontal personal agents and whether they can be used as both commercial platforms and advertising platforms. “In some regions, the popularity of these horizontal personal agents may be far slower than expected.”

Justin Post (Bank of America Research): In addition to agency economics, whether AI agents will prioritize direct-run websites over OTA; and the impact of Q4 cost increases and 2027 investments on growth algorithms.

Fogel: “It's interesting that I've heard this statement, but I've also seen other scenarios — OTAs are actually getting a bigger advantage right now.” The key is to provide more value to travelers and agents, so agents are willing to access, display, and trade.

Steenbergen: The investment philosophy is “making a disciplined investment for the long term while protecting immediate profitability.” We don't have a lot of capitalization, and the cost base is completely realistic, including equity incentives — I don't understand why anyone would want to remove this; this is the real cost of running a business.” Commenting on 2027 is too early, and will commence when the results for the fourth quarter of February next year; however, the pre-order data for the first half of 2027 “looks healthy”, boosting confidence in next year.

Ronald Josey (Citi Research): What metrics to look for from the Lola and Vayo trial fields, what to learn and then promote; and the value of the booking supply when agents become new entrants.

Fogel: The two new brands are positioned to explore the travel business with different agency-based methods to observe what customers like, what converts well, and what has signed up a lot — for example, “most users are willing to authorize the import of email reservation history, which is a very good idea” on Vayo. Details of specific indicators are not disclosed; once a certain ability is successful, it can be ported to major platforms such as Booking, Agoda, and Priceline.

Fogel (supplementary supply value): It's hard to explain why Booking has more exposure on these platforms. The company has some theories — related to the level, reliability, and model judgment of content value; there is no “manual on how to show up in LLM or personal agents”, which ultimately provides more value, so that in agent-agent conversations, horizontal agents recognize that recommended booking is the best choice for travelers.

Steenbergen: Brand trustworthiness is an important factor — horizontal agents can obtain “instant credibility” by linking with the Booking brand; they tend to work with aggregators because it is too expensive to directly connect 4.7 million listings and obtain real-time prices and property status data, and there is still a possibility of establishing a monetization model with the company in the future.

Fogel: If these horizontal agents move towards some kind of monetization model, “it's not fundamentally different from what we've been doing for the past 20 years, and we'll be in an even better position”.

Benjamin Black (Deutsche Bank Research Division): Genius accounts for an increasing share of transactions, how to deepen loyalty without relying more on vendor funding discounts; and whether the operational efficiency brought about by AI may outperform the “8%/15%” long-term EPS algorithm.

Fogel: It will improve Genius and make it “a better project in the future”; Genius currently works well because partners fund and get good ROI, and travelers get benefits; there are many other forms of value that can be explored in the future, some with almost zero marginal cost to partners; no specific plans are disclosed, but it is well aware that it is necessary to continue to provide differentiated value to form a “more loyal, more consumer, and more beneficial” flywheel for “travelers.”

Steenbergen: It's too early to judge how AI will change long-term algorithms; but AI has “moved from experiments to company-wide measurable operational value” — traveler experience side (engagement, conversion, booking value, incrementality), customer service side (single booking cost drops, satisfaction remains high), engineering side (cost of a single merge request has dropped significantly); personnel growth has been extremely low in the past two years, and is mainly concentrated in capacity centers at low cost locations such as India.

Meredith Prichard Jensen (HSBC Research): Are changes in the distribution landscape opening new doors for collaboration between Booking and hotel suppliers on loyalty programs and guest identification — for example, hotel projects booked through the Muse channel, and Booking as the connecting party.

Fogel: The more nodes the network has, the value is growing exponentially; the company is exploring loyalty program connections with hotels, various travel players, and even partners outside the travel ecosystem to get more value; this is an important part of ensuring Genius maintains one of the most valuable loyalty programs; it is not convenient to disclose exactly what it is doing, but it knows that it is necessary to continue to provide more value to both sides of the network.

Concluding remarks

Fogel: Thank you to our employees for their dedication and efficient execution, our shareholders for their support, and our travelers and partners who trust us every day.” We — all of us — are incredibly excited about the role AI will play in travel. I am very encouraged by our progress on various strategic priorities. We will navigate the recent uncertainties with the same discipline and invest prudently in the opportunities that lie ahead. Our talent and the capacity we continue to build give us confidence to continue to play a leading role in shaping the future of travel and create lasting value for our shareholders. Thank you so much and good night.”

The moderator announced the end of the meeting.