The direct answer: In Q2 2026, Tripadvisor’s hotel comparison and display segment (“Hotels and Other”) fell 21% year over year to $163.3 million — from $208 million a year earlier — a decline the company itself ties to Google AI Overviews eroding the organic search traffic that hotel advertisers pay to reach (Tripadvisor Q2 2026 earnings, August 6, 2026). In the same quarter, rival metasearch Trivago grew revenue 21% — its sixth consecutive double-digit quarter — and raised full-year guidance (Trivago, August 4, 2026). Same demand environment, same product category, opposite outcomes. That divergence is the single most useful data point available for planning your 2027 hotel marketing budget, because it shows the metasearch category is not dying uniformly. It is splitting in two: comparison traffic that depends on classic blue-link SEO is collapsing, while comparison traffic reachable through direct marketing and AI-native surfaces is holding or growing. If your 2027 budget still treats “metasearch” as one line item, you are allocating against a category that no longer exists in that form.
What actually happened in Q2 2026
Tripadvisor’s quarter was rough by any standard. Total revenue fell 7% to $441.9 million, net income dropped 38% to $22.8 million, results missed analyst forecasts by 12.6%, and the stock fell roughly 25% on the day (earnings reported August 6, 2026). The detail that matters to hotels is the segment split:
- Hotels and Other: $163.3M, down 21% — this is the hotel price-comparison and metasearch advertising business. Q3 2026 guidance says it will fall another 20–23%.
- Experiences (Viator + branded tours): $278.6M, up 3% — with Viator itself growing in the mid-teens, now listing 400,000+ bookable experiences.
CEO Matt Goldberg named the cause on the earnings call: “ongoing declines in flyby visitors to our site due to the changing search landscape and the rise of AI overviews.” Marketing spend rose 4% to $215.4 million even as revenue fell — Tripadvisor is now paying more to acquire fewer visitors. When a middleman’s traffic economics break, the advertising rates they charge you eventually follow. Some of that spend is advertising hotels to an audience that is shrinking underneath it.
The strategic backdrop reinforces the retreat: Tripadvisor agreed in June 2026 to sell TheFork to American Express for $700 million, signed a cooperation agreement with activist investor Starboard in March 2026 (four of ten board seats), and management told investors “all options are open” on the remaining businesses — including, per Skift’s analysis, the hotel metasearch unit itself. The platform whose reviews feed AI answers may be restructuring away from the hotel-comparison business hotels used to advertise on.
Meanwhile Trivago — the same category, Europe-centric — delivered 21% revenue growth in Q2 2026 after 15% growth in Q1 2026, and is fighting Google in court over hotel search dominance. Demand for comparing hotel prices did not disappear. It moved.
The mechanism: AI answers absorb the comparison query
Three 2026 data points explain the split better than any narrative:
- AI Overviews now appear on roughly 48% of US search queries as of March 2026, up from 31% a year earlier, with measured click-through rate reductions of 30–50% on informational queries where they appear. More than 58% of US queries now end without any external click (Hotel Online, June 2, 2026). “Best hotel in [city]” is exactly the informational-comparison query class being answered in place.
- In Google AI Mode, citation is not traffic. A 2026 study of 4,000 hotel-related queries by researcher Nicolas Sitter found Tripadvisor cited in 86% of hotel searches — but nearly all those links route to Google Business Profiles rather than to Tripadvisor’s own site. A separate February 2026 analysis of eight cities found roughly 79% of clickable hotel links inside AI Mode responses route first to a Google Business Profile, and only about 17% go directly to a hotel’s own website (The Hotel Blueprint, August 31, 2026).
- Comparison is migrating inside assistants. Kayak’s AI Mode — launched October 2025 — evolved into a global “Ask AI” on its homepage by April 2026, and Kayak’s What the Future report (September 3, 2026) confirms travelers now use KAYAK.ai to surface deals, not just draft itineraries. Phocuswright (March 2026) found 44% of travelers would book directly inside an AI platform.
Read together: the traveler still compares. But the comparison increasingly happens inside an AI answer or an AI-native interface, and the click that survives usually lands on your Google Business Profile — not on Tripadvisor’s page where your metasearch bid was waiting, and not on your website unless your AI visibility is deliberately built. We covered the routing mechanics in depth in Google AI Mode and hotels: where the clicks actually go; this quarter’s earnings make the commercial consequence impossible to ignore.
What the numbers mean for your 2027 budget
| Channel | What the 2026 data shows | Budget implication for 2027 |
|---|---|---|
| Tripadvisor hotel metasearch / display | Hotels & Other revenue −21% in Q2 2026 ($208M → $163.3M); Q3 guidance −20% to −23%; owner restructuring around the unit | Audit any spend that depends on Tripadvisor’s organic traffic; keep review operations (reviews still power AI citations), renegotiate rates, expect inventory quality to drift |
| Trivago | +21% Q2 2026, sixth consecutive double-digit quarter, FY guidance raised to mid-teens | Comparison demand is alive where it is marketed directly; keep campaigns where CPC-to-booking math still clears your direct-booking value |
| Google AI Overviews / AI Mode | ~48% of US queries show AIOs (Mar 2026); CTR −30–50% on affected queries; ~79% of AI Mode hotel links route to Google Business Profiles | Treat your Business Profile as a primary conversion surface: rates, photos, attributes, Q&A, Hotel Ads linking — not an SEO footnote |
| Kayak Ask AI / KAYAK.ai | Ask AI global on homepage since April 2026; WTF 2026 report (Sept 2026) shows deal-seekers comparing inside AI interfaces | Small early-test budget; AI-native placements are cheap attention while everyone else waits |
| Direct + AI visibility | 44% of travelers would book inside an AI platform (Phocuswright, Mar 2026); direct bookings carry zero commission | Fund GEO work that makes you citable in AI answers — the only channel where you are the destination, not a bid |
The reallocation playbook: five moves
1. Split “metasearch” into two lines before you cut anything. Line A: campaigns whose economics depend on a middleman’s organic Google traffic (Tripadvisor-dependent placements). Line B: campaigns priced on direct-response inventory you can still measure (Trivago where it works, Google Hotel Ads). Line A is a melting ice cube — harvest it while the ROAS lasts, but do not renew it at 2025 rates. Line B still earns its place quarter by quarter.
2. Promote your Google Business Profile to a revenue channel. If ~79% of AI Mode hotel clicks land on your profile, that page is effectively your homepage for AI-era comparison. Fresh rates, complete attributes, current photos, answered questions, and a connected Hotel Ads feed do more for the 2027 funnel than another branded-campaign CPC increase. This is also where our AI visibility work for hotels starts: you cannot fix routing you have not measured.
3. Move 10–15% of metasearch spend into AI visibility measurement and content. Not 100% — a wholesale exit would be as reflexive as staying put. A measured reallocation, spent on a travel AI audit baseline, entity and schema fixes, and answer-shaped content, buys you presence in the interfaces absorbing the comparison query. Phocuswright’s 56% AI adoption figure means your next guest has likely already asked ChatGPT or Gemini about you — the only question is whether the answer included you.
4. Make the direct booking path machine-usable. Every AI-routed traveler you do win should land on a booking flow that is fast, structured, and parity-safe. If assistants or AI Mode send traffic to your profile or site and the path to book is worse than the OTA’s, you paid twice: once for visibility, once in commission. We break down the measurement side in measuring AI-driven direct bookings.
5. Watch the OTAs’ own AI moves — and the regulatory floor. Google has already announced third-party rates for hotel ads will be unavailable after September 30, 2026 (O’Rourke Hospitality, September 24, 2026), and AI Max / Performance Max now power ads shown inside AI Mode (Google Marketing Live, July 2026). The comparison layer is being rebuilt while you read this. A quarterly prompt audit — the same set of 25–50 buying prompts, run across ChatGPT, Gemini, Perplexity and AI Mode — is now a budget-control instrument, not a curiosity. Our travel GEO agency runs exactly this cadence for hotels and DMCs, and pricing is fixed so the audit pays for itself before any retainer conversation.
What not to do
Do not read TripAdvisor’s −21% as “metasearch is dead” and cut comparison spend wholesale — Trivago’s +21% in the same quarter is the counter-proof. Do not assume Tripadvisor’s decline makes reviews irrelevant either: the Sitter study shows Tripadvisor is cited in 86% of AI Mode hotel searches. Your review profile now works for the AI answer rather than for Tripadvisor’s page — which is precisely why review operations deserve budget even as metasearch placements there get renegotiated. And do not wait for the 2027 planning cycle to fix the Google Business Profile; the routing data says that asset is already doing the job your metasearch bids used to do.
FAQ
Is hotel metasearch dead in 2027? No — it is split. Tripadvisor’s hotel comparison segment fell 21% in Q2 2026 with guidance for a further 20–23% decline in Q3, while Trivago grew 21% in the same quarter (its sixth consecutive double-digit quarter). Comparison demand persists; it is migrating away from SEO-dependent surfaces toward direct-marketing channels and AI-native interfaces like Kayak’s Ask AI.
Should our hotel stop advertising on Tripadvisor? Not automatically. Split the decision: placements whose ROI depends on Tripadvisor’s shrinking organic traffic are a harvest-and-exit candidate; your review operations on the platform still matter because Tripadvisor is cited in 86% of AI Mode hotel searches (Sitter, 2026) — you just no longer need to pay for their traffic to benefit from their citation. Renegotiate, measure last-click value quarterly, and let the ROAS decide the wind-down pace.
What is replacing metasearch for hotel comparison? Three things, in order of current volume: Google Business Profiles (receiving ~79% of AI Mode hotel clicks), AI assistants and chat interfaces where comparison happens conversationally (ChatGPT, Gemini, Perplexity, Kayak Ask AI), and surviving direct-response metasearch where the CPC math still works (Trivago in Europe). None of them replace the others; all three now need deliberate management.
How much of our 2027 budget should move to AI visibility? A defensible starting point is 10–15% of current metasearch and brand-CPC spend, redirected to a baseline audit, entity/schema fixes, answer-shaped content, and the Google Business Profile upgrade. Scale it on evidence: prompt-set citation tracking and AI-referral conversion, not platform reach estimates.
What is the first practical step? Run a baseline audit of where you currently appear — and where competitors appear — across the 25–50 prompts your guests actually use, plus a Google Business Profile readiness check. That baseline, repeated quarterly, becomes the scoreboard every 2027 budget decision points at. You can start it with a travel AI audit.
Related reading
- Google AI Mode and hotels: where the clicks actually go
- Measuring AI-driven direct bookings
- The OTA AI counterattack: what booking platforms are doing about direct bookings
The comparison query did not die in 2026. It moved house — quietly, one AI answer at a time. Your 2027 budget is the document that decides whether you follow it.