Global lodging and online travel: World Cup-driven lodging strength fades, while OTA traffic increasingly diverges

Overview

This global lodging and online-travel Q3 update argues that an exceptionally strong, World Cup-supported July is giving way to more normal hotel demand. Bernstein uses traffic data to forecast OTA room nights and regional RevPAR run rates to assess hotel operators, while also highlighting generative AI as an early but growing source of travel discovery traffic.

Core views

Bernstein's central hotel-industry conclusion is that July's US lodging strength was event-driven and is now normalizing. The World Cup knockout rounds helped lift US RevPAR 8.3% year on year in July, with Luxury RevPAR up 17.7%. By August, US RevPAR growth had slowed to 2%, and luxury growth was at its weakest since January. The report expects September broadly to resemble August, implying healthy but modestly decelerating Q3 RevPAR growth versus Q2 for most operators. Marriott and Hyatt are expected to lead the hotel groups because their greater US luxury exposure captured more of the World Cup benefit; Hilton was running behind its approximately 4% RevPAR guidance in the first two months of the quarter. Regional trends are uneven. Europe continued mid-single-digit RevPAR growth in Q3, while APAC excluding China grew at roughly 9% in July and August. In contrast, China saw year-on-year RevPAR declines for four consecutive months, Americas excluding the US contracted year on year in July and August, and US momentum slowed after the tournament ended. Middle East and Africa returned to growth, with Egypt and Jordan offsetting weaker Gulf performance. The report also notes that a Labor Day timing shift distorted early-September US weekly data; its 28-day moving average nevertheless indicates a clear slowdown since mid-July, particularly in luxury. For online travel agencies, Bernstein finds widening dispersion between app and web traffic. App engagement growth decelerated for Airbnb, Booking and Expedia versus Q2, while web traffic was generally more resilient. Airbnb is described as the cleanest setup: Q3-to-date web traffic grew 17.7% year on year and app downloads grew about 30%, supporting Bernstein's estimate of 152.0 million Q3 room nights, or 13.8% growth and 2.1% above the 149.0 million consensus estimate. The model does not yet include an effect from elevated cancellations or a potential positive effect from RNPL at Airbnb. For Booking, Bernstein compares Q3-to-date traffic with Q3 2025 and changes in traffic since Q2 guidance, then averages the two approaches. Its forecast is 343.6 million room nights, equivalent to 6.4% growth and 1.4% above the 338.8 million consensus estimate. Although Booking's web growth had slowed to about 8% from low-teens growth in the first half, the report says web traffic improved through September and that Booking historically tends to beat its guidance range by about 2% when post-guidance traffic trends do not change. Expedia is the relative concern. Total app MAUs fell 11% year on year, with declines across every brand and particularly weak trends at Vrbo. Web traffic rebounded and reached mid-teens growth by September, but Bernstein considers the app weakness more consequential for its B2C room-night outlook. It estimates 112.0 million Q3 room nights, or 3.5% growth, marginally below the 112.1 million consensus estimate. Tripadvisor also continued to experience traffic declines across web and app, although Viator's web traffic returned to positive growth in June and improved through September. The report identifies AI-driven travel discovery as a developing structural issue rather than a material current earnings driver. AI referrals to OTAs rose from about 0.5% of total traffic in June to roughly 0.7-0.8% in August. Hotel websites received around 1.0-1.2% of traffic from AI platforms for Hilton, Hyatt and IHG, and approximately 1.7% for Marriott. The report links the recent reversal of the prior pattern of app traffic outgrowing web traffic to potentially greater agentic engagement. New tools such as Muse and Instinct can search hotels, compare prices and direct users toward OTA or hotel websites, which Bernstein expects could increasingly shape traffic distribution. On longer-term lodging supply, Bernstein says the global hotel pipeline has recovered meaningfully through Q3-to-date after an April pullback tied to global macroeconomic disruption. Rooms under construction improved during the quarter, although the percentage of the overall pipeline under construction declined as the pipeline expanded faster. The relative size of the pipeline remains below mid-2023 lows, STR's declared net unit growth is about 1.7%, branded pipeline share remains within a 78-80% range, and the large asset-light groups have helped drive year-to-date pipeline strength. Marriott has the largest global pipeline share, Hilton's share is increasing, IHG's is flat, and Hyatt's share is stable, though Bernstein cautions that STR tends to understate Hyatt's pipeline substantially.

Analysis framework

Bernstein combines third-party web and app traffic data with historical traffic-to-room-night relationships to build quarterly OTA room-night predictors. It compares estimates with consensus and company guidance, and uses July-August RevPAR data, regional performance, weekly 28-day moving averages and hotel-pipeline data to assess hotel operating momentum and supply conditions.

Methodology notes

• Other

  Traffic-based quarterly room-night predictor

  Bernstein uses quarter-to-date web traffic, app engagement, downloads and MAUs, together with observed historical correlations between traffic and room nights, to estimate Q3 room-night volumes for Airbnb, Booking and Expedia.

• Industry AnalysisSupply-demand framework

  RevPAR and hotel-pipeline tracking

  The report reads hotel demand through regional and chain-scale RevPAR growth, then assesses future supply through global pipeline, rooms-under-construction and net-unit-growth data.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Airbnb (ABNB)

  Bernstein expects Q3 room nights to outperform consensus, supported by strong web traffic.

  Strengths

  Q3-to-date web traffic grew 17.7% year on year; app downloads grew about 30%.

  Weaknesses

  App engagement growth has continued to slow.

  Comparison

  Bernstein forecasts a 2.1% room-night beat versus consensus, ahead of Booking and Expedia.

  Risks

  The estimate excludes any impact from elevated cancellations and any potential RNPL benefit.

• Booking Holdings (BKNG)

  Bernstein expects a modest Q3 room-night beat versus consensus.

  Strengths

  Traffic trends improved after guidance, including a strong start to September.

  Weaknesses

  Web traffic growth slowed to about 8%, while app engagement growth was about 5% and downloads were broadly flat year on year.

  Comparison

  Estimated Q3 room nights are 1.4% above consensus, below Airbnb's 2.1% projected beat.

  Risks

  Room-night performance relative to guidance is sensitive to changes in traffic after guidance.

• Expedia (EXPE)

  Weak app traffic leads Bernstein to expect a slight room-night miss versus consensus.

  Strengths

  Web traffic rebounded and reached mid-teens growth by September; B2B growth remained robust.

  Weaknesses

  Total app MAUs fell 11% year on year, with declines across all brands.

  Comparison

  Estimated Q3 room nights are 0.1% below consensus, versus projected beats for Airbnb and Booking.

  Risks

  Persistent app-engagement pressure could outweigh web and B2B strength.

• TripAdvisor (TRIP)

  Bernstein rates TripAdvisor Outperform despite weak traffic across its core channels.

  Strengths

  Viator web traffic returned to positive growth in June and improved through September.

  Weaknesses

  Tripadvisor traffic continued to decline across both web and app channels.

  Risks

  Continued traffic declines in the core Tripadvisor business.

• Marriott (MAR)

  Luxury exposure supported stronger Q3 RevPAR run-rate performance.

  Strengths

  Greater US luxury exposure; largest share of the global hotel pipeline.

  Weaknesses

  July-August run rates are likely to decline as the World Cup-supported July comparison rolls off.

  Comparison

  Bernstein says Marriott and Hyatt outpaced guidance in the first two months, while Hilton trailed its approximately 4% guidance.

  Risks

  Normalization in US luxury RevPAR after the World Cup.

• Hyatt (H)

  Luxury exposure supported stronger Q3 RevPAR run-rate performance.

  Strengths

  Greater US luxury exposure; Bernstein notes STR may materially understate its pipeline.

  Weaknesses

  STR shows Hyatt's global pipeline share as stable.

  Comparison

  Alongside Marriott, Hyatt outpaced guidance in the first two months of Q3.

  Risks

  Normalization in US luxury RevPAR after the World Cup.

• Hilton (HLT)

  Hotel operator with improving global pipeline share but softer near-term RevPAR run rate.

  Strengths

  Its global pipeline share is steadily increasing.

  Weaknesses

  Running behind approximately 4% Q3 RevPAR guidance in the first two months.

  Comparison

  Lagged Marriott and Hyatt on July-August RevPAR run-rate performance.

  Risks

  AI traffic remains a small but evolving source of hotel-website demand.

• InterContinental Hotels Group (IHG.LN)

  Covered global hotel operator.

  Strengths

  Its website received approximately 1.0-1.2% of traffic from AI platforms in August.

  Weaknesses

  Its share of the global pipeline was described as flat.

  Comparison

  Marriott has the largest pipeline share and Hilton's share is increasing.

Key data

• US RevPAR growth8.3% in July; 2% in AugustYear-on-year growth slowed after World Cup tailwinds faded.
• US Luxury RevPAR growth17.7% in JulyWorld Cup benefit was most pronounced in the luxury segment.
• Airbnb Q3 room nights152.0 millionBernstein estimate versus 149.0 million consensus, 2.1% above consensus; implied 13.8% growth.
• Booking Q3 room nights343.6 millionBernstein estimate versus 338.8 million consensus, 1.4% above consensus; implied 6.4% growth.
• Expedia Q3 room nights112.0 millionBernstein estimate versus 112.1 million consensus, 0.1% below consensus; implied 3.5% growth.
• Expedia app MAUs-11%Year-on-year decline in Q3-to-date total app MAUs.
• Airbnb web traffic growth+17.7%Year-on-year Q3-to-date growth supporting its room-night estimate.
• AI referral traffic to OTAs~0.5% in June to ~0.7-0.ахыс8% in AugustShare of total traffic; still small despite rapid growth.
• APAC excluding China RevPAR growth~9% in July and AugustYear-on-year growth cited as resilient relative to weaker China and Americas excluding the US.
• STR declared net unit growth~1.7%Global hotel supply-growth indicator.

Impact & implications

Bernstein sees near-term hotel earnings support shifting away from a broad event-driven demand boost and toward operator-specific exposure, with Marriott and Hyatt benefiting from luxury mix. For OTAs, traffic divergence supports relative room-night outperformance at Airbnb and Booking, while Expedia's weak app engagement is a drag despite stronger web and B2B trends. AI referrals are not yet large enough to drive aggregate volumes, but may progressively alter how travel demand is routed between OTAs and hotel websites.