Google Ads for travel agencies & OTAs 2026: the complete vertical playbook

Travel advertising on Google has its own infrastructure — Hotel Ads, Things to do, travel feeds, and a dedicated Hotel campaign type — sitting alongside the standard Search, Performance Max, and Demand Gen surfaces every other vertical uses. That extra surface area exists because travel intent is unusually structured: people search a destination, compare prices and availability across providers, and then book days or weeks later, often after multiple sessions. No other consumer category combines such high intent with such a long, multi-touch path to purchase.

This guide is written for travel marketers — at OTAs, tour operators, activity providers, hotel groups, and independent agencies — who need a campaign architecture that respects how travel actually converts. We will cover seasonality and booking windows, the Google travel surfaces, where Performance Max helps and where it quietly destroys margin, long-window attribution, dynamic pricing in ads, targeting that scales, and the uncomfortable reality of competing against advertisers with effectively unlimited budgets.

Why travel is the hardest vertical to run on Google Ads

Three structural features make travel uniquely difficult, and each one breaks a default Google Ads assumption:

• The conversion path is long and non-linear. A leisure traveler researching a trip might search five times across two weeks, on three devices, before booking — and then book 30-60 days before travel. Google Ads' default 30-day click window and last-click instincts both truncate this. You will systematically under-credit upper-funnel destination research and over-credit the final branded click, then defund the very campaigns that started the journey.
• CPCs are brutal because the auction is stacked with giants. Booking Holdings and Expedia Group spend billions on performance marketing annually, and they bid on nearly every commercial travel query. Generic destination terms routinely clear €1.50-4.50 CPC, and high-intent dated booking terms reach €3-9. A small operator bidding head terms is buying clicks at a loss unless conversion rates and margins are exceptional.
• Inventory is perishable and priced dynamically. An empty seat or unsold room tonight is worth zero tomorrow, and prices change hourly. Ads that show a stale price, or that send a user to a sold-out date, waste spend and damage trust. Travel demands feed freshness and availability automation that most retail accounts never have to think about.
• Demand is shock-prone. Weather, currency swings, airline strikes, public-health news, and viral destinations can move demand 50% in a week. Static bids and budgets cannot keep pace; the account has to be watched and adjusted continuously.

Layer on top of this the volatility of demand — weather, events, currency swings, airline strikes, viral destinations — and you have a vertical where set-and-forget bidding fails fast. The operators who win treat the account as a living system: feeds refreshing, attribution windows tuned to real lag, bids scheduled against research demand, and Performance Max kept on a tight leash. The rest of this guide builds that system piece by piece. For the seasonality mechanics in depth, our seasonality budget guide and machine-learning seasonal bid adjustments go deeper than we can here.

Travel demand seasonality and the booking-window problem

Travel seasonality is two overlapping curves, and confusing them is the most common and most expensive mistake in the vertical.

The travel-date curve is when people physically travel: summer for beaches, December for ski and holidays, shoulder seasons for city breaks. The search-demand curve is when people research and book — and it leads the travel-date curve by the length of the booking window. Booking-window data varies by segment, but typical leisure patterns look like this:

The practical workflow: pull at least two years of your own booking data, compute the distribution of (travel date minus booking date), and identify your true search-demand peaks. Then schedule budget and bid increases against those peaks. If 40% of your summer bookings happen in March, your March budget should reflect summer demand even though no one is traveling yet.

Google Trends is a useful confirmation layer for destination-level search interest, and it often reveals demand you cannot serve from historical bookings alone (a destination going viral, a new route opening). But your own booking-window distribution must lead — Trends shows interest, not your conversion lag.

Two more seasonality tactics matter:

• Seasonality adjustments in Smart Bidding for known short spikes — a 48-hour flash sale, a public-holiday weekend — where you expect a conversion-rate jump too brief for the algorithm to learn organically.
• Separate early-booker and last-minute campaigns rather than blending them, because they have opposite economics: early-bookers need inspiration and price-lock messaging at moderate bids, while last-minute travelers convert at much higher rates and justify aggressive CPCs inside a tight window.

The Google travel surfaces: Things to do, Hotel Ads, Travel feeds

Beyond standard campaigns, Google offers travel-specific inventory that high-intent travel shoppers see directly in Search and Maps. Using these surfaces is what separates a travel-native account from a generic one.

Hotel Ads (the Hotel campaign type). Hotel Ads place your property and live rate inside Google's hotel search and Maps experience, where users compare prices across providers. To participate you need a hotel feed: static property data plus an availability and price feed delivered through the Hotel Prices API or an approved connectivity partner (the integration most small and mid-size hotels use). Bidding can run on commission (pay per stay or per conversion) or CPC, and the dedicated Hotel campaign type lets you set bids by check-in day, length of stay, device, and audience. Because the user is comparing rates side by side, rate parity and feed freshness are decisive — an out-of-date rate either loses the click or, worse, wins it and then disappoints on the landing page.

Google Things to do. This surface shows bookable tickets, tours, and activities for attractions directly in Search and Maps. Operators connect through a Things to do feed or a connectivity partner, and listings can appear as free booking links (organic) or paid placements. For attraction operators, museums, tour companies, and experience providers, this is some of the highest-intent inventory available — the user is already researching a specific attraction and sees a bookable option in place. It pairs naturally with in-destination, mobile-first bidding because much activity demand is same-day.

Travel feeds for dynamic ads. Beyond the two surfaces above, a dynamic remarketing feed powers personalized display and Performance Max ads that rebuild themselves with the exact destination, property, or trip a user viewed, plus live pricing. This is where feed quality compounds: one well-maintained feed can fuel Hotel Ads, dynamic remarketing, and Performance Max prospecting simultaneously.

Feed quality is the make-or-break discipline across all three surfaces. The minimum bar for any travel feed:

• Unique, stable IDs per property, activity, or trip so reporting and remarketing stay consistent.
• Live price and availability, refreshed at least daily and ideally near-real-time.
• Destination and geo data so the right item serves to the right intent.
• Image URLs and rich titles that survive moderation and render well across surfaces.
• Prompt removal of sold-out inventory so ads never point at a dead landing page.

The connectivity question deserves a note: most operators who are not large enough to build direct API integrations work with an approved connectivity partner that manages the feed plumbing for Hotel Ads and Things to do. Budget for that relationship — the alternative is engineering time you probably do not have. For the accommodation-specific deep dive, see our hotels and hospitality Google Ads guide.

Performance Max for travel: when it helps, when it hurts

Performance Max is genuinely powerful for travel — and genuinely dangerous if you deploy it the way Google's onboarding flow nudges you to. The difference is whether you ring-fence it.

Where it helps. Performance Max shines at two travel jobs:

• Dynamic remarketing. Fed with your travel feed and seeded with audience signals (cart abandoners, recent destination searchers, past guests), it reassembles ads with the exact trip a user looked at and chases them across YouTube, Gmail, Discover, Display, and Search.
• Discovery-stage prospecting. For inspiration-driven travel, a themed asset group per destination can surface your offers to people who have not yet entered a specific search.

For both jobs, the cross-surface reach and feed-driven personalization are hard to replicate manually.

Where it hurts. Left unconstrained, Performance Max does three damaging things in travel accounts. It cannibalizes branded search, quietly serving on people typing your brand name and claiming those cheap conversions as its own — which inflates its reported ROAS and makes it look better than it is. It over-credits low-intent placements because the booking window is long and the model attributes generously. And it hides where spend goes, giving you minimal placement and search-term visibility precisely in the vertical where you most need to know whether you are paying for incremental demand or just intercepting your own.

The 2026 architecture that resolves this:

Concretely, the guardrails that keep Performance Max honest in a travel account:

• Brand exclusions on every Performance Max campaign, so it cannot dress up branded traffic as net-new revenue.
• Account-level negative keywords and excluded placements to keep it off junk inventory and irrelevant queries.
• A budget cap of 20-30% of total spend until you have proven incrementality, ideally via a geo holdout test.
• Real audience signals (cart abandoners, recent searchers, past guests) rather than letting it prospect blind.
• Skepticism toward its reported ROAS until you have validated it against incremental lift, not just last-click credit.

Our deeper analysis of how unconstrained Performance Max erodes accounts lives in why Performance Max destroys 30% of accounts and the complete Performance Max guide.

Attribution across 30-60 day booking windows

The long booking window is travel's defining measurement problem, and getting it wrong silently misallocates your entire budget.

Extend the conversion window. Google Ads defaults to a 30-day click conversion window, but leisure travel routinely books 30-60 days after the first click — and long-haul stretches to 120. If your window is shorter than your booking lag, you simply will not see the conversions your upper-funnel campaigns produced, and Smart Bidding will defund them. Set the click conversion window to 60-90 days for most travel accounts, longer if your data shows it.

Switch to data-driven attribution. Last-click is actively harmful in a multi-session, multi-device vertical. Data-driven attribution distributes credit across the destination-research, comparison, and booking touches, which keeps prospecting funded instead of starving it to feed branded last clicks.

Import the value that actually happens, not the value booked. Two refinements separate good travel measurement from great:

• Stitch the real booking to the click. If bookings finalize in a separate booking engine or PMS rather than on the click-tracked page, use offline conversion import (joining the GCLID captured at click to the eventual booking) so Google optimizes toward real revenue, not a deposit or a thank-you page.
• Optimize for stays, not bookings. Travel has meaningful cancellation rates, so import net-of-cancellation revenue with a second upload 30-45 days after booking. If you optimize toward gross bookings, Smart Bidding will happily chase segments that book and then refund.

For the mechanics of conversion tracking and offline import, see our conversion tracking guide. And because travel buyers cross devices and sometimes channels before booking, cross-channel attribution is worth understanding once your single-channel measurement is solid.

A final caution on ROAS targets: report and optimize against contribution margin, not revenue. An OTA earning a 12% commission needs roughly an 8x revenue ROAS just to break even before fixed costs; a direct operator at 50% margin breaks even near 2x. Setting Target ROAS in revenue terms without anchoring to margin is how travel accounts hit their ROAS goal and lose money simultaneously.

Dynamic pricing and availability in ads

Travel prices move constantly and inventory expires, so your ads must reflect live data or they actively work against you. This is a feed-quality discipline as much as a campaign one.

Live price in the ad. Showing a current, accurate price does two things: it pre-qualifies the click (price-sensitive shoppers who would bounce never click) and it preserves trust through to the landing page. The mechanics depend on the surface:

• Hotel Ads pulls rates from your price and availability feed.
• Dynamic remarketing pulls from your remarketing feed.
• Search surfaces prices via price assets and dynamic ad customizers.

In all cases the feed must refresh at least daily and ideally near-real-time; a price that is stale by even a few hours can trigger disapprovals or, worse, a bait-and-switch experience when the landing-page rate differs.

Availability-aware serving. An ad for a sold-out date or property is pure waste, and it frustrates users. Use feed availability flags to drive automation: pause or bid down items as they sell out, and drop unavailable rates from the Hotel feed promptly so they stop serving. Where possible, redirect that demand — if a property sells out, keep capturing the destination intent by routing budget to comparable available alternatives rather than going dark.

Ad customizers for Search. For text ads, dynamic ad customizers let you inject live values — price, dates, "X rooms left", countdown to a sale end — without rebuilding ads. Used well, they make Search ads feel as current as the feed-driven surfaces. (For a full treatment of dynamic ad mechanics across formats, our dynamic creative optimization tutorial covers customizers and dynamic remarketing setup end to end.)

Sold-out as an opportunity, not just a problem. Sophisticated operators turn scarcity into a tactic: when popular inventory sells out, they lean into urgency messaging on the remaining availability and on comparable alternatives ("Lisbon selling fast — 3 similar stays available"). Scarcity is one of the few honest urgency levers in travel, and feed-driven availability data is what lets you wield it accurately rather than crying wolf.

The connective tissue across all of this is feed governance: a single source of truth that updates prices and availability frequently, validates required fields, and surfaces disapprovals fast. Travel accounts live or die on feed health long before they live or die on bid strategy.

Destination plus intent targeting that actually scales

Travel keyword strategy is a matrix of destination times intent times booking window, and structuring it deliberately is what makes spend efficient at scale.

The destination-by-intent matrix. Think in two axes. The destination axis runs from broad ('city breaks Europe') to specific ('hotel Alfama Lisbon'). The intent axis runs from inspiration through comparison to booking:

• Inspiration ('best time to visit Lisbon') — cheap, high-volume, low conversion. Useful for upper-funnel reach and remarketing-list building, but ruinous if bid like booking terms.
• Comparison ('Lisbon vs Porto', 'best Lisbon neighborhoods') — mid-intent, mid-cost. Capture and remarket; rarely your last click.
• Booking ('book Lisbon hotel January') — expensive and converting. Fund hard, and only here should you tolerate top-of-auction CPCs.

Each cell deserves different bids, creative, and landing pages. Mixing them into one ad group lets the cheap inspiration clicks drain budget meant for booking intent.

Layer the booking window. On top of the matrix, separate early-booker and last-minute ad groups as discussed earlier, with date and urgency modifiers ("last minute", "tonight", "this weekend", "2026") splitting traffic into distinct intent pools you can bid independently.

Match types and negatives in 2026. With broad match now leaning heavily on Smart Bidding signals, broad match plus Target ROAS can work for destination discovery — but only behind a disciplined negative-keyword program. Travel attracts enormous irrelevant volume: jobs ("travel agent jobs"), DIY and free ("free things to do"), news, visa and government queries, and informational research with no booking intent. A comprehensive shared negative list is non-negotiable, and you should mine search-term reports weekly because new junk appears constantly.

Geo targeting both ends of the trip. Travel is unusual in caring about two locations: where the traveler is (origin) and where they want to go (destination). Handle each differently:

• Origin targeting for most prospecting — target by where the traveler lives and let the keyword carry the destination.
• Destination targeting for in-destination demand — activities, restaurants, last-minute rooms — with a tight radius and mobile-first bids, because the highest-intent activity searches come from people already there with their phones out.

Audience layers complete the picture: apply in-market travel audiences and your own remarketing lists as observation on Search to inform bid adjustments, and as targeting on Performance Max. The compounding asset here is your first-party data — searchers, viewers, and especially past guests. For building durable first-party audiences in a cookieless world, see Customer Match and first-party data.

Competing with Booking, Expedia and the OTA giants

If you sell travel on Google, you share every commercial auction with Booking Holdings and Expedia Group — companies whose annual performance-marketing budgets exceed the lifetime revenue of most independent operators. Pretending you can out-bid them is the fastest way to burn a budget. Winning means refusing to fight on their terms. The five-front strategy in brief:

• Concede most head terms — fight where your economics work, not where theirs do.
• Own the long tail — niche, specific, intent-rich queries the OTAs under-cover.
• Sell the direct-booking value — give people a concrete reason to leave the OTA.
• Defend and exploit your brand — your cheapest, highest-ROAS inventory.
• Move faster on volatility — speed is a real edge against large, slow incumbents.

Concede the head terms, mostly. Generic, high-volume terms ("hotels in Paris", "cheap flights") are where the OTAs' budgets and conversion-rate optimization are strongest and where your CPCs are least defensible. Either skip them or test them only behind a hard ROAS cap and tight match control. The math rarely works for a smaller advertiser, and that is fine — there is more profitable demand elsewhere.

Own the long tail. OTA coverage thins out on specific, niche, and intent-rich queries: "boutique eco-lodge Costa Rica family", "dog-friendly cottage Cornwall short break", "small-group photography tour Iceland". These convert well, cost far less per click, and are where a specialist operator's actual differentiation lives. Many independents profitably concentrate the majority of spend on long-tail plus branded plus remarketing and treat head terms as an occasional, capped experiment.

Give people a reason to leave the OTA. A user clicking your ad instead of Booking's needs a concrete reason: a genuine price advantage, direct-booking perks (free cancellation, room upgrades, loyalty points), unique inventory, or expertise the OTA cannot match. Put that value proposition in the ad copy and prove it on the landing page. Direct-booking messaging is your structural counter to the OTA's distribution advantage.

Defend and exploit your brand. Run branded search to protect your own name (OTAs and competitors will bid on it), and treat branded plus remarketing plus past-guest Customer Match as your cheapest, highest-ROAS inventory. These audiences cost a fraction of cold prospecting and convert far better — fund them first, every time, before you spend a euro fighting giants on open terms.

Be faster on volatility. The OTAs are large and, in places, slow. Independents can move quickly on a viral destination, a sudden route, a weather-driven demand shift, or a competitor's stock-out. Speed on volatility is a real edge for a nimble operator who is watching demand daily.

Put together, the budget allocation that works for most independents looks roughly like this:

• 40-50% long-tail and niche destination/intent — your most defensible, highest-margin demand.
• 15-20% branded search — protect the name competitors and OTAs bid on.
• 15-20% remarketing and past-guest Customer Match — the cheapest, highest-ROAS inventory you own.
• 10-15% Performance Max prospecting — ring-fenced, capped, brand-excluded.
• 5-10% capped head-term experiments — only behind a hard ROAS ceiling.

A free, structured audit is the fastest way to see where an OTA-heavy account is leaking budget — typically unconstrained Performance Max cannibalizing brand, a too-short conversion window hiding the booking lag, and head-term spend that never had a chance. SteerAds runs a free 14-day audit on Google and Microsoft Ads that surfaces exactly these travel-specific failure modes.

Sources

• developers.google.com/hotels — Google Hotel Ads and Hotel Prices API documentation
• developers.google.com/travel/things-to-do — Google Things to do connectivity documentation
• thinkwithgoogle.com — Think with Google travel research and consumer-journey data
• skift.com/research — Skift travel-industry research and OTA market analysis
• support.google.com/google-ads — Performance Max and brand exclusions documentation

FAQ