[BKNG] Booking Holdings: Q3 2026 Earnings and the Long-Haul Drag on Room Nights

Booking Holdings runs the world's largest online travel platform through five consumer brands, Booking.com, Priceline, Agoda, KAYAK and OpenTable, and earns most of its revenue from accommodation commissions and the transaction revenue it keeps after collecting payment from travelers[1]. Booking Holdings Q3 2026 earnings are due on 2026-10-27, when the company holds its call on the third quarter of 2026, ending September 30, 2026[2]. In the latest reported period, the second quarter of 2026, room nights rose 5.3% to 325 million[3], gross bookings rose 9.0% to $50.96 billion[4], revenue rose 8.1% to $7.35 billion[5], adjusted EBITDA reached $2.648 billion at a 36.0% margin[6], and adjusted EPS rose 15% to $2.54[7]. On August 4, management guided third-quarter room nights up 3% to 5% and gross bookings, revenue and adjusted EBITDA each up 4% to 6%, with constant-currency revenue up 5% to 7% and currency trimming about 1 percentage point[8]. As of September 29, the analyst consensus compiled by Drillr stood at third-quarter revenue of $9.544 billion (19 analysts, range $9.444 billion to $9.733 billion), EBITDA of $4.762 billion, and post-split EPS of $4.48 (18 analysts, range $4.29 to $4.62)[9].

Three things matter most in this report. First, whether room nights stay inside the 3% to 5% range: second-quarter growth leaned on domestic travel while international room nights grew only slightly[10], and the guidance assumes higher airfares, reduced long-haul capacity and soft long-haul demand persist through the whole third quarter[8], so the results will show whether the Middle East conflict's drag on lodging is narrowing or spreading. Second, whether revenue keeps pace with gross bookings: merchant bookings reached 73% of second-quarter gross bookings[11], yet revenue as a share of gross bookings slipped to 14.4% from 14.5% a year earlier[5], and with identical growth ranges guided for revenue and gross bookings, the third quarter is the first test of the company's booking-versus-travel timing explanation. Third, whether marketing costs are rising: marketing expense was 4.7% of gross bookings in the second quarter, up from 4.6%[12], and the company said in its quarterly filing that free search traffic will keep declining in the short to medium term and could push up paid marketing spend[13], so the third-quarter marketing ratio and adjusted EBITDA are the most direct test of whether AI search pressure has reached the financials.

Company Background and Business Structure

Booking Holdings is a lodging-centered, multi-brand online travel platform that was formerly known as Priceline, and it completed a 25-for-1 stock split on April 2, 2026, so all per-share figures since then are shown on a post-split basis[14]. Booking.com focuses on accommodation, Priceline mainly serves North America, Agoda mainly serves Asia, KAYAK runs metasearch referrals, and OpenTable sells restaurant reservations and restaurant-management subscriptions[1]. In 2025 the platform booked 1.235 billion room nights, up 8.0%, and 68 million airline tickets, up 36.6%[15], and about 89% of revenue came from online accommodation reservations[16]. The company's long-term strategy is the "Connected Trip," which links lodging, flights, car rental and attractions in a single booking; in the second quarter these transactions grew more than twice as fast as Booking.com's total transactions and made up a low double-digit percentage of them[10].

The company reports a single segment and splits revenue into merchant, agency, and advertising and other, with merchant revenue now clearly dominant. Merchant revenue is the commission and transaction net revenue Booking keeps after collecting from the traveler and paying the hotel, plus payment-related revenue such as card rebates, processing fees and travel insurance; agency revenue is the commission Booking collects after a traveler pays the hotel directly, almost all from Booking.com lodging; advertising and other revenue comes mainly from KAYAK referral and advertising fees and OpenTable reservation and subscription fees[1]. In 2025, merchant revenue rose 25.5% to $17.755 billion, agency revenue fell 6.5% to $7.968 billion, and advertising and other revenue rose 11.3% to $1.194 billion, or roughly 66%, 30% and 4% of the total[16]. In the second quarter of 2026, merchant revenue grew 15.0% to $5.127 billion, agency revenue fell 6.9% to $1.903 billion, and advertising and other revenue grew 8.1% to $322 million[5].

Booking does not own room inventory; it takes a commission share of what hotels, home-rental hosts, airlines and other suppliers earn, and this delivery model shapes both its seasonality and its cash timing. Gross bookings are recorded when a traveler books, revenue is recognized at check-in, and marketing expense falls in the quarter of booking, so the third quarter, which contains the European and North American summer check-in peak, is the most profitable quarter of the year and the first quarter the least[17]. Under the merchant model the traveler pays first and Booking settles with the hotel after the stay, which builds a large deferred merchant bookings balance of $10.121 billion at the end of June 2026[18]; the cost is that payment processing, chargebacks and fraud losses grow with merchant transactions and are recorded in personnel and in sales and other expenses[11].

Financial History and Current Position

Booking's revenue recovered quickly after the pandemic and kept reaching new highs in 2023 and 2024. Revenue rose from $10.958 billion in 2021 to $21.365 billion in 2023, when operating income was $5.835 billion and net income $4.289 billion[19]. Revenue then reached $23.739 billion in 2024[16], with net income of $5.882 billion[20].

In 2025 the company moved up another step in scale, although one-off impairment and currency losses distorted reported profit. Full-year revenue rose 13.4% to $26.917 billion[16], and gross bookings rose 12% to $186.107 billion, or about 10% in constant currency[21]. Marketing expense rose 12.5% to $8.186 billion, about 4.4% of gross bookings; operating income was $8.825 billion and net income $5.404 billion, below 2024[20]. Operating income included a $457 million goodwill and intangible impairment for KAYAK taken in the third quarter[22]. Operating cash flow was $9.409 billion and capital expenditure $322 million, for free cash flow of about $9.087 billion, while buybacks totaled $6.438 billion and dividends $1.248 billion[23].

Growth slowed noticeably in the first two quarters of 2026 but still beat the company's guidance. First-quarter room nights rose 6% to 338 million, with the company estimating a drag of about 2 percentage points from the Middle East conflict, and gross bookings rose 15% to about $53.8 billion, or about 8% in constant currency[24]; revenue grew 16% and the adjusted EBITDA margin was 23.3%[14]; adjusted EPS was $1.14 on a post-split basis[24]. Second-quarter operating income was $2.500 billion and net income $1.950 billion[25], adjusted EBITDA grew 9%, operating cash flow was $3.719 billion and free cash flow $3.643 billion[6], and the company said every key second-quarter metric beat the top of its guidance[10].

The balance sheet shows that capital returns have outrun first-half free cash flow and are partly funded with new debt. In the first half of 2026 Booking paid $7.758 billion for buybacks and $664 million in dividends, issued $2.975 billion of long-term debt, and generated $6.934 billion of operating cash flow[26]. At the end of June it held $17.7 billion of cash and investments, about $14.7 billion of it in international subsidiaries; senior notes had $20.3 billion of principal, including $19.1 billion in euro-denominated debt; and stockholders' equity stood at a deficit of $10.783 billion[18]. Remaining buyback authorization was $14.5 billion at the end of June[12].

Operating Model

Revenue can be written as gross bookings times revenue as a share of gross bookings; lodging demand drives the first term, while the merchant shift, flight mix and currency drive the second. Lodging bookings roughly equal room nights times average daily rate, and second-quarter gross bookings growth of 9.0% came from about 5% room-night growth, roughly 2% higher constant-currency room rates, about a 1% currency tailwind and growth in flight bookings[4]. Revenue is recognized at check-in, and revenue was 14.4% of gross bookings in the second quarter and 12.3% in the first half[5]; subtracting the first half from the 2025 full year implies second-half 2025 revenue of about $15.357 billion on gross bookings of about $92.701 billion, a ratio near 16.6%, which shows how heavily summer check-ins weight the second half[16]. The 2025 increase in the revenue-to-bookings ratio came mainly from higher payment-related revenue, partly offset by a larger mix of flights, which carry high ticket values but low commission rates[21]; a longer booking window pushes revenue later, and currency moves change reported growth directly because Booking reports in dollars while earning most of its revenue in euros and other currencies[17].

Operating income equals revenue minus marketing, sales and other expenses and a set of largely fixed costs, and marketing is the biggest swing factor for margins. Booking has no cost of goods; most marketing spend is performance marketing paid to search engines such as Google, affiliate channels, metasearch and social media, and second-quarter marketing expense of $2.371 billion equaled 32.2% of revenue and 4.7% of gross bookings[13]. Because marketing is expensed at booking while revenue is recognized at check-in, faster booking growth compresses margins first, and when demand slows marketing adjusts before revenue does[17]. Two other levers shape margins: the gap between payment revenue and payment costs from the merchant shift, which the company says was positive over the trailing twelve months[11], and the transformation program's cut to fixed costs, where the company raised expected annual run-rate savings to about $650 million and expects most of the amount above the roughly $550 million level reached at the end of 2025 to arrive in 2027[27]. Net income also swings with euro-debt currency effects, from a $961 million loss in the second quarter of 2025 to a $195 million gain in the second quarter of 2026, and these items do not reflect operating change[6].

Deferred merchant bookings set the rhythm of cash flow: the company collects heavily in the first half and releases that cash in the second half as travelers check in. In the first half of 2026, deferred merchant bookings and other current liabilities rose by $4.207 billion[26], lifting the deferred merchant bookings balance from $5.270 billion at the start of the year to $10.121 billion at the end of June[18], and this prepaid cash flows out gradually in the second half as travelers stay and Booking pays hotels. The third quarter's operating cash flow is therefore usually lower than the second quarter's, and whether buybacks can continue at the first-half pace depends on how much cash the company is willing to use and how much new debt it adds.

Industry and Competitive Position

Booking's advantages come from scale, a long tail of supply and its payment capabilities. In the second quarter, alternative accommodation room nights at Booking.com, such as vacation homes and apartments, grew 4% and made up 37% of its total room nights, flat with a year earlier[10]; the merchant model lets the company handle multi-currency collection, refunds and fraud for many independent hotels and small hosts. Geographically, the business is centered on Europe and Asia, and the United States is a market it has pushed harder in recent years, with U.S. room-night growth accelerating for a fourth straight quarter to the low teens in the first quarter of 2026[24].

Competition comes from several directions, and generative AI has made the contest for traffic entry points less predictable. The annual report lists competitors including other online travel companies, airlines, hotels and rental car firms selling direct, and companies such as Airbnb and Expedia-owned Vrbo that compete directly in lodging; it specifically notes that Google links travel search to Maps and its Gemini generative AI, and that other large technology platforms and AI-native companies are building assistants and agents that can search, compare and book directly, which could reduce visits to dedicated travel platforms[28]. Booking also depends on traffic from third-party platforms such as Google, and competition for placement in AI-generated results could raise costs per click and hurt marketing efficiency[29]. Because the company does not disclose brand or regional revenue or a quarterly direct-channel share, the competitive comparison here rests on the company's own qualitative statements and room-night mix and cannot be carried into a line-by-line financial comparison with rivals.

Regulation is a clear disadvantage for Booking in Europe. The company is designated a "gatekeeper" under the EU Digital Markets Act, and Booking.com has removed price-parity clauses in the European Economic Area as required[30]. Spain's competition authority penalized Booking.com in 2024 over parity clauses and ranking rules, and although the company has appealed it carried a $472 million liability at the end of June; the Swiss Price Surveillance Office ordered a cut in the average commission charged to hotels in Switzerland, an order suspended during the appeal that would take effect for three years after a final loss[31].

Core Debates

Domestic travel carried second-quarter room nights while the Middle East conflict held back long-haul trips. Can third-quarter room nights stay within the 3%-5% guidance?

This is the most important question because lodging generates about 89% of revenue and the third quarter is the year's highest-revenue and highest-profit quarter[16]. Second-quarter room nights rose 5.3% to 325 million and revenue rose 8.1%, both above the top of guidance[10]; but the growth came from domestic travel, and airline ticket growth slowed from 16.0% in the first half to 3.7% in the second quarter[3]. The company guided third-quarter room nights up 3% to 5% and revenue up 4% to 6%, assuming long-haul capacity and fare pressure last through the end of the quarter[8]; it also cut full-year gross bookings guidance because of slower flight growth while leaving its lodging outlook largely unchanged[32].

The evidence for resilient demand is that the conflict's impact is fading and domestic demand remains strong. First-quarter room nights grew 6%[14], and second-quarter room nights grew 5.3%, with global domestic room nights up high single digits and strong domestic demand in the United States and Asia[10]; the company said the conflict's impact began normalizing in June and continued to normalize in July, and full-year guidance remains in line with its original constant-currency expectations[33]. The evidence against is that the conflict hit the second quarter harder than the first[33], elevated March cancellations reduced second-quarter revenue[34], and currency has turned from a first-half tailwind into a headwind of about 1 percentage point in the third quarter[32]. An alternative reading is that the second-quarter beat came mainly from a one-time substitution of domestic trips for long-haul travel plus higher room rates, and if domestic demand fades after the summer while long-haul travel does not recover, room-night growth could fall below 3%.

The financial transmission of this debate is direct. The conflict grounds routes, higher fuel costs push up airfares and long-haul capacity shrinks, so long-haul international travel and inbound Middle East demand fall while domestic travel only partly fills the gap; room-night growth and average rates set lodging gross bookings, which turn into revenue in the quarter of check-in. Because marketing is spent at booking, a demand slowdown squeezes margins before it hits revenue. What remains unresolved is whether domestic substitution is a durable change in demand mix or a one-time summer shift.

In the third-quarter report, watch whether room nights grow at least 3%, whether revenue reaches the 4% to 6% range and how the company explains the currency drag, as well as constant-currency gross bookings growth, fourth-quarter guidance, and the trend in airline tickets and international and long-haul room nights. If room nights grow less than 3% and the company blames weaker long-haul or domestic demand, or if it cuts full-year gross bookings or revenue guidance, the current view that lodging demand is resilient would be falsified.

Booking now collects payment on most of its bookings. Can the extra payment revenue keep covering its costs so that revenue keeps pace with gross bookings?

Merchant bookings made up 73% of total gross bookings in the second quarter, up from 69% a year earlier[11], and the merchant model is the foundation that lets Booking.com control payment and push the Connected Trip and its Genius loyalty program. How much revenue each dollar of bookings produces determines how fast revenue can still grow once room-night growth drops to the mid single digits. Revenue was 14.4% of gross bookings in the second quarter, 0.1 percentage point lower than a year earlier, which the company attributed to the timing of bookings versus travel[5]; third-quarter guidance calls for the same 4% to 6% growth in revenue and gross bookings[8], making it the first quarter to test that explanation.

The supporting evidence is that the merchant shift has indeed let revenue outgrow bookings, and payment costs remain manageable. The merchant mix rose from 63% in 2024 to 70% in 2025[16], and revenue as a share of bookings rose because of higher payment-related revenue[21]; the company says incremental payment revenue exceeded incremental variable costs over the trailing twelve months[11]; and sales and other expenses fell to 12.8% of second-quarter revenue from 13.2%, mainly thanks to efficiencies in third-party customer service costs[35]. The contrary evidence is that revenue as a share of bookings fell year over year in both the second quarter and the first half[5], first-half merchant transaction costs rose by $149 million and grew faster than total revenue[35], and a rising flight mix itself pulls the ratio down[21]. An alternative reading is that the migration dividend is fading: the revenue growth above bookings in recent years came from a one-time migration rather than durable payment monetization.

The transmission chain runs as follows: Booking.com shifts agency transactions to the merchant model, the merchant mix rises, and by collecting payment the company adds card rebates, processing fees and other payment revenue that lift revenue as a share of bookings; at the same time payment processing, chargeback and fraud costs flow into sales and other expenses, and the net effect shows up in the gap between revenue growth and bookings growth and in operating margin. On second-quarter gross bookings of $50.957 billion, each 0.1 percentage point decline in the revenue-to-bookings ratio removes about $50 million of quarterly revenue[4]. What remains unresolved is whether incremental payment revenue can keep covering rising transaction costs once the merchant mix nears its ceiling.

In the third-quarter report, watch the gap between revenue growth and gross bookings growth, whether the merchant share of bookings keeps rising year over year, and sales and other expenses as a share of revenue along with the company's explanation of payment costs and customer-service efficiency. If revenue growth trails bookings growth by more than 1 percentage point, or if sales and other expenses rise as a share of revenue, the view that payment revenue still covers its costs would be falsified.

Google's AI search is eroding free traffic. How much more marketing spend will Booking need to hold on to its bookings?

Marketing is Booking's largest cost, at $8.186 billion in 2025 or about 30% of revenue[20], and most of it goes to Google and other search and performance channels[13]. The company's annual report also lists AI assistants and agents that bypass online travel platforms and reduce direct traffic as a competitive risk[28], which is exactly the market's recent concern. The third-quarter marketing ratio and adjusted EBITDA are the most direct measures of whether that concern has reached the financials; the second-quarter baseline is marketing at 4.7% of gross bookings (4.6% a year earlier) and adjusted EBITDA of $2.648 billion, up 9%, at a 36.0% margin[12].

The evidence that the impact is limited is that AI channels are still tiny and the company's own channels are getting stronger. Management said traffic from large language models accounts for significantly less than 1% of room nights, with no material change in recent quarters[33]; the direct channel's share of room nights held at a mid-fifties percentage over the trailing twelve months[13], and the app share rose from a mid-fifties to a high-fifties percentage[11]; Genius level 2 and 3 members account for half of room nights, and AI voice support now handles most eligible inbound calls, cutting customer-service cost per booking by a double-digit rate year over year[36]. The contrary evidence is that the company said in its quarterly filing that free search traffic will keep declining in the short to medium term and could raise paid marketing spend[13], the second-quarter marketing ratio rose 0.1 percentage point year over year[12], and KAYAK has already taken a $457 million impairment because of expected increases in customer acquisition costs[22]. An alternative reading is that margin expansion came mainly from one-time savings in the transformation program and customer service, and once those savings slow, rising channel costs will show up directly in margins.

The financial transmission runs along three lines. Display changes such as Google's AI overviews reduce free search traffic and intensify bidding for ad slots in AI results, so more bookings depend on paid channels and marketing cost per booking rises; rising direct, app and Genius shares offset this, and together they set marketing as a share of gross bookings. Fewer metasearch referrals directly pressure advertising and other revenue, where KAYAK sits, and that line was $322 million in the second quarter, up 8.1%[5]. The transformation program lowers fixed costs and supports the adjusted EBITDA margin[27]. On second-quarter bookings, each 0.1 percentage point rise in the marketing ratio adds about $50 million of quarterly expense, roughly 2% of second-quarter adjusted EBITDA[6]. What remains unresolved is whether AI agents end up as just another referral channel or systematically raise customer acquisition costs.

In the third-quarter report, watch whether the year-over-year change in marketing as a share of gross bookings exceeds 0.2 percentage point, whether adjusted EBITDA reaches the 4% to 6% growth guidance[8], what the company now says about its direct-channel share and LLM traffic share, and whether advertising and other revenue turns to decline. If the marketing ratio rises more than 0.2 percentage point year over year, or adjusted EBITDA grows less than 4%, the view that AI diversion has limited impact would be falsified.

Risks and Falsifiers

The first risk is capital structure and cash timing: buybacks have exceeded free cash flow and are partly debt-funded, while merchant prepayments unwind seasonally with third-quarter check-ins. In the first half of 2026 buybacks of $7.758 billion exceeded roughly $6.75 billion of free cash flow, and the company issued $2.975 billion of new debt[26]; at the end of June senior notes carried $20.3 billion of principal, with currency moves on the $19.1 billion of euro debt flowing through net income, stockholders' equity stood at a deficit of $10.783 billion, and the $10.121 billion of deferred merchant bookings will unwind with second-half stays[18]. If third-quarter free cash flow still covers that quarter's buybacks and dividends and the company issues no new debt, this concern would be falsified.

The second risk is European regulation and commission limits, which act directly on commission rates and revenue. The Spanish case already carries a $472 million liability, and if the Swiss commission-cut order takes effect after a final ruling it would lower average commissions on Swiss hotels for three years[31]; the Digital Markets Act has already forced the company to drop parity clauses in the European Economic Area[30]. If the appeals keep their suspensive effect and no new commission limit or fine appears during the quarter, this risk will not reach the financials this quarter.

The third risk is a renewed escalation of the Middle East conflict or a sustained contraction in long-haul airline capacity that depresses international travel and room nights. The third quarter is the year's highest-revenue quarter, and second-half 2025 revenue was about $15.357 billion[16]; at the second-quarter revenue scale, each 1 percentage point of lost room-night growth removes about $70 million of quarterly revenue, and because marketing is spent at booking a demand slowdown squeezes margins first[5]. If third-quarter room nights grow at least 5% and the company confirms long-haul demand improved from the second quarter, this risk would be falsified.

The fourth risk is a fading merchant-migration dividend: once the merchant mix nears its ceiling, incremental payment revenue slows while chargeback, fraud and payment-processing costs keep rising with volume[11]. On second-quarter gross bookings of $50.957 billion, each 0.1 percentage point drop in the revenue-to-bookings ratio removes about $50 million of quarterly revenue; on second-quarter revenue, each 0.5 percentage point rise in sales and other expenses as a share of revenue adds about $37 million of quarterly expense[35]. If third-quarter revenue growth is no lower than bookings growth and sales and other expenses fall as a share of revenue, this risk would be falsified.

The fifth risk is AI diversion and search change: AI assistants and agents from Google, OpenAI and other platforms could capture travel demand or raise click costs, while free search traffic keeps declining[34]. On second-quarter bookings, each 0.1 percentage point rise in the marketing ratio adds about $50 million of quarterly expense, roughly 2% of second-quarter adjusted EBITDA, and KAYAK's metasearch revenue would come under pressure at the same time[22]. If the third-quarter marketing ratio is flat or down year over year, the direct-channel share holds, and the company confirms LLM traffic is still below 1%, this risk would be falsified.

What to Watch Next

• Room nights: the second-quarter baseline is 325 million, up 5.3%[3]; watch whether third-quarter growth lands inside the 3% to 5% guidance[8]. Growth below 3% blamed on weaker long-haul or domestic demand weakens the resilience view, while growth of at least 5% with better long-haul demand falsifies the conflict risk.
• Revenue, currency and full-year guidance: second-quarter revenue was $7.352 billion, up 8.1%[5]; watch whether growth reaches 4% to 6% and how the roughly 1-point currency drag is explained. A cut to full-year bookings or revenue guidance weakens the current view.
• Constant-currency bookings, flights and international room nights: second-quarter gross bookings were $50.957 billion, up 9.0%, and airline tickets grew 3.7%[4]; watch fourth-quarter guidance and flight and long-haul trends. A recovery in long-haul travel and flights would confirm the conflict drag is narrowing.
• Revenue growth minus bookings growth: revenue was 14.4% of bookings in the second quarter versus 14.5% a year earlier[5]; with both guided at 4% to 6%, revenue trailing bookings by more than 1 point falsifies the payment-coverage view.
• Merchant share and sales and other expense ratio: merchant mix was 73% versus 69%[11], and the expense ratio was 12.8% versus 13.2%[35]; a year-over-year rise in the expense ratio falsifies the view.
• Marketing as a share of bookings: 4.7% versus 4.6%[12]; a rise of more than 0.2 point falsifies the limited-AI-impact view.
• Adjusted EBITDA: $2.648 billion, up 9%, at a 36.0% margin[6]; growth below 4% falsifies the view.
• Direct channel, LLM traffic and advertising and other revenue: direct share is in the mid-fifties and LLM traffic is significantly below 1%[33], while advertising and other revenue was $322 million; a decline in advertising and other revenue weakens the view.
• Cash and capital returns: first-half buybacks of $7.758 billion exceeded roughly $6.75 billion of free cash flow[26]; free cash flow covering third-quarter buybacks and dividends without new debt would falsify the concern.

Conclusion

Three variables drive Booking: room nights and room rates set gross bookings, the merchant shift decides how much revenue each dollar of bookings becomes, and marketing efficiency decides how much of that revenue turns into profit. The company remains financially strong, with 2025 revenue of $26.917 billion and free cash flow of about $9.087 billion[23], and its second-quarter adjusted EBITDA margin still widened to 36.0%[6]; but room-night growth has slowed from 8.0% in 2025 to 5.3% in the second quarter, third-quarter guidance points lower to 3% to 5%[8], and first-half buybacks exceeded free cash flow and leaned partly on new debt. The key unresolved relationship is whether payment revenue and savings in customer service and fixed costs can keep offsetting rising marketing costs while long-haul travel stays under pressure and growth slows.

After the second-quarter results, CNBC's Isabella Gao argued in a September 9 analysis that treating Booking as a loser from the AI platform shift does not fit the operating data: LLM traffic is still below 1% of room nights, and while AI agents can plan a trip, executing bookings at global scale requires payments, compliance, live inventory and customer service, so Google's and OpenAI's early agent products hand checkout back to the online travel agency, with Booking a launch partner in both[37]. The piece also concedes that AI is genuinely eroding thin, top-of-funnel metasearch businesses such as KAYAK, argues that AI has so far lowered Booking's customer-service costs, and notes that a sizable part of the first-half slowdown in growth reflected dollar-based currency effects. This view points the same way as the company's own statements on the marketing-efficiency debate and offers a currency-based reading of the lodging-demand debate, but it depends on the premise that AI agents are just a new referral channel that will not materially raise acquisition costs. Between the second-quarter report and the end of September, this was the only data-backed outside analysis available, with other coverage mostly recapping share-price moves and news, and no equally argued bearish analysis appeared, so it should not be read as a shared market judgment.

Going forward, the current view that domestic demand provides a floor while payment monetization and cost savings offset channel pressure would be strengthened if third-quarter room nights land within 3% to 5% or near the top of that range, revenue growth is no lower than bookings growth, marketing as a share of bookings is flat or down year over year, adjusted EBITDA grows 4% to 6%, and free cash flow covers that quarter's buybacks and dividends. That view would be clearly weakened if room nights fall below 3% growth, revenue growth trails bookings by a wide margin, the marketing ratio rises more than 0.2 point year over year, or the company cuts full-year guidance and keeps funding buybacks with new debt.

Sources

[1] BKNG 10-K filed 2026-02-18 · business model and revenue types · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[2] Drillr earnings calendar (updated 2026-09-29) · BKNG 2026-10-27 call · 2026-09-29 · Drillr earnings calendar

[3] BKNG 10-Q filed 2026-08-04 · Q2 2026 room nights, rental car days and flight tickets · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[4] BKNG 10-Q filed 2026-08-04 · Q2 2026 gross bookings and drivers · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[5] BKNG 10-Q filed 2026-08-04 · Q2 2026 revenues by type · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[6] BKNG 8-K filed 2026-08-04 · adjusted EBITDA and free cash flow reconciliation · 2026-08-04 · 8-K · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=8-K

[7] BKNG 8-K filed 2026-08-04 · Q2 2026 results headline · 2026-08-04 · 8-K · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=8-K

[8] BKNG 8-K filed 2026-08-04 · Q3 and FY2026 guidance · 2026-08-04 · Booking Holdings · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=8-K

[9] Drillr analyst_financial_estimates (updated 2026-09-29) · BKNG quarter ending 2026-09-30 · 2026-09-29 · Drillr analyst_financial_estimates · https://gateway.drillr.ai/mcp/private

[10] BKNG Q2 2026 earnings call 2026-08-04 · segment performance · 2026-08-04 · earnings-call · https://gateway.drillr.ai/mcp/private

[11] BKNG 10-Q filed 2026-08-04 · merchant mix and payment economics · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[12] BKNG 8-K filed 2026-08-04 · Q2 2026 margins, marketing and capital return · 2026-08-04 · 8-K · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=8-K

[13] BKNG 10-Q filed 2026-08-04 · Q2 2026 marketing expenses and direct channel · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[14] BKNG 8-K filed 2026-04-28 · Q1 2026 results and stock split · 2026-04-28 · 8-K · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=8-K

[15] BKNG 10-K filed 2026-02-18 · FY2025 operating metrics · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[16] BKNG 10-K filed 2026-02-18 · FY2025 revenues and gross bookings by type · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[17] BKNG 10-K filed 2026-02-18 · seasonality and booking-to-travel timing · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[18] BKNG 10-Q filed 2026-08-04 · June 30 2026 balance sheet, cash and debt · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[19] BKNG 10-K filed 2024-02-22 · FY2021-FY2023 statements of operations · 2024-02-22 · 10-K · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-K

[20] BKNG 10-K filed 2026-02-18 · FY2025 operating expenses and income · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[21] BKNG 10-K filed 2026-02-18 · FY2025 gross bookings and take-rate drivers · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[22] BKNG 10-K filed 2026-02-18 · KAYAK goodwill and intangible impairment · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[23] BKNG 10-K filed 2026-02-18 · FY2025 cash flow and capital return · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[24] BKNG Q1 2026 earnings call 2026-04-28 · segment performance · 2026-04-28 · earnings-call · https://gateway.drillr.ai/mcp/private

[25] BKNG 10-Q filed 2026-08-04 · Q2 2026 statement of operations · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[26] BKNG 10-Q filed 2026-08-04 · H1 2026 cash flow statement · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[27] BKNG 10-Q filed 2026-08-04 · transformation program update · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[28] BKNG 10-K filed 2026-02-18 · competition and Gen AI · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[29] BKNG 10-K filed 2026-02-18 · marketing efficiency and platform dependence · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[30] BKNG 10-K filed 2026-02-18 · DMA gatekeeper and regulation · 2026-02-18 · 10-K · https://www.sec.gov/Archives/edgar/data/1075531/000107553126000009/

[31] BKNG 10-Q filed 2026-08-04 · CNMC and Swiss commission proceedings · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[32] BKNG Q2 2026 earnings call 2026-08-04 · Q3 and full-year guidance detail · 2026-08-04 · earnings-call · https://gateway.drillr.ai/mcp/private

[33] BKNG Q2 2026 earnings call 2026-08-04 · Q&A on LLM traffic, flights and Q3 · 2026-08-04 · earnings-call · https://gateway.drillr.ai/mcp/private

[34] BKNG Q2 2026 earnings call 2026-08-04 · stated risks · 2026-08-04 · earnings-call · https://gateway.drillr.ai/mcp/private

[35] BKNG 10-Q filed 2026-08-04 · Q2 2026 sales and other expenses · 2026-08-04 · 10-Q · https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0001075531&type=10-Q

[36] BKNG Q2 2026 earnings call 2026-08-04 · Genius, AI and transformation · 2026-08-04 · earnings-call · https://gateway.drillr.ai/mcp/private

[37] CNBC 2026-09-09 · This battered travel stock is seen as an AI victim · 2026-09-09 · CNBC · https://www.cnbc.com/2026/09/09/this-battered-travel-stock-is-seen-as-an-ai-victim-but-the-data-suggests-otherwise.html