Say your Google Ads dashboard reports a cost per lead of 38 euros and everything looks healthy. Your CRM, three months later, tells a different story: the enquiries that turned into paid deposits came from the campaigns with the worst-looking numbers.
The gap is structural. In travel, the click and the money are months apart. Someone searches from a sofa in Rotterdam in January, downloads an itinerary, asks for a quote, pays a deposit in March and travels in September. Google sees the click. Your bank sees the deposit. Unless you connect the two, the bidding algorithm optimises towards whichever keyword produces the cheapest form fill, and in travel those are reliably the ones with the least money behind them.
So here is the playbook: structuring campaigns around trip intent, matching keywords to lead time, feeding booking value back into the bidding, pacing budget against the booking curve, and measuring it all when your sales cycle outlives the default conversion window.
Structure the account around trips, not destinations
Most travel accounts are built like the website: one campaign per destination. It looks tidy and hides everything that matters. Inside a single Peru campaign sits someone pricing a two-week honeymoon, someone wanting a four-day Inca Trail add-on, and a student hunting the cheapest way to Cusco. Same country, three different businesses, one bid strategy.
Structure instead by what you sell and how it earns. The unit that deserves its own campaign is the trip type: its own price band, margin, lead time and landing page. A fixed group departure is not a tailor-made itinerary that takes six hours to quote. A workable skeleton:
- Brand on exact and phrase in its own campaign, so you can see what defending your own name costs.
- One campaign per product line with real margin, split by source market where seasonality differs.
- Destination plus qualifier, for searches naming a place and a shape of trip: self-drive, honeymoon, 10 days, with kids.
- Competitor and category terms, capped and judged on booked margin, since these clicks are expensive and slow.
- Search remarketing, bidding up for people who read an itinerary in full and left without enquiring.
Match the keyword to the trip type and the lead time
Travel searches sit at very different distances from a deposit. Best time to visit Namibia is somebody eight months out with no supplier in mind. Namibia self-drive itinerary 10 days is someone shaping a trip, three to five months out. Namibia tailor-made safari operator is choosing a company to trust, often weeks from paying.
Do not run them against the same target. Buying-intent terms carry your revenue number. Shaping terms should carry a content offer, a route map or sample itinerary, feeding an email sequence built for a consideration window measured in months rather than days. Inspiration terms belong to SEO and email until you can hold a lead that long without flinching at the cost.
Negatives do more work here than almost anywhere, because the vocabulary of travel is shared by people with no money to spend with you:
- Jobs and training: guide jobs, tour leader vacancies, internships, courses.
- Do-it-yourself and free: free itinerary, how to plan it yourself, budget backpacking, hostels.
- Adjacent products you do not sell: flights only, visas, insurance, car hire, currency.
- Wrong audience shapes: school trips, volunteering, relocation, digital nomad.
- Destination-side searches such as things to do today, if you sell to visitors from source markets.
Why clicks and cost per lead break down in travel
A single account-wide cost-per-lead target instructs Google to find you the cheapest possible enquiry. In travel, cheap enquiries are abundant and mostly poor: price-checkers, people who wanted a flight, people with 600 euros for a 6,000 euro trip. The algorithm is doing exactly what you asked. You asked for the wrong thing.
Take an illustration with invented, round numbers. Two campaigns, 2,500 euros of media each. Campaign A sells family camping in France: 100 enquiries at 25 euros, one in five books, so 20 bookings averaging 1,400 euros. At a 15 percent margin that is 210 euros each, or 4,200 euros of gross margin. Campaign B sells tailor-made Namibia: 20 enquiries at 125 euros, one in four books because those searchers have already decided they want an operator, so 5 bookings averaging 9,000 euros for two travellers. At an 18 percent margin that is 1,620 euros each, or 8,100 euros.
Same spend. Campaign B costs five times more per enquiry and returns nearly twice the margin. An account-wide 40 euro target would have switched off the better business by Friday. The numbers are invented; the shape is not. Once trip value and close rate vary this much between lines, a single cost-per-acquisition target is a blunt instrument. Run the same arithmetic on your own product lines before you argue with it.
Feed booking value back into the bidding
Google optimises towards the outcomes it is told about. If the last thing it hears is form submitted, form fills are what it keeps buying. And in travel a form fill is a long way from a sale: it is the opening of a conversation that runs through a quote, two or three itinerary revisions, a conversation about single supplements and finally a deposit. So capture the Google Click ID when the enquiry arrives, keep it on the record through your pipeline, and send the later stages back with values attached.
Google Ads Help sets out the constraints in its offline conversion imports FAQs: the click ID is retained for 90 days, conversions carrying personally identifiable information for enhanced conversions can be uploaded so long as they are no older than 63 days, and uploads should happen at least daily. Its value-based bidding documentation adds that you should upload values for three weeks, or one to two conversion cycles, whichever is longer, before activating value-based bidding.
One piece of housekeeping before you build anything. Google Ads Help states that from 15 June 2026, offline conversion imports and enhanced conversions for leads uploads migrated to the Data Manager API and are blocked in the Google Ads API, and that operators not already importing offline conversions should start with enhanced conversions for leads. If your developer wired the upload against the old endpoint, check that before you trust a single number the account reports.
Then the travel-specific problem. If your median time from click to deposit is 120 days, the deposit never matches back inside the 90-day click ID window, so import an earlier stage instead. Quote sent usually lands in time. Value it at your average booked margin for that line multiplied by that stage's close rate. Staying with the invented Namibia figures above, quotes closing at one in four against 1,620 euros of margin would make a sent quote worth about 405 euros. Use your own numbers, not those. And check the conversion window while you are in there: Google Ads Help gives a 30-day click-through default and allows up to 90 days depending on the conversion source, which matters when your cycle runs four months.
This is a CRM job more than an ads job. Whether you run a travel CRM such as OpenVoy or a tidy pipeline in whatever you already have, the click ID must survive from enquiry to deposit and stage changes must export daily. Send margin, not revenue: a 9,000 euro trip with flights bought at cost is not a 9,000 euro win, and bidding towards the gross price will push you towards the flight-heavy itineraries that earn you least.
Search, Performance Max and Demand Gen: what each one is for
Search is where intent already exists, and where most travel companies should spend most of their money most of the year. It is the only channel where somebody tells you the destination, the shape of the trip and often the timing, unprompted, in their own words.
Performance Max needs strong conversion signal and a real catalogue behind it. Google Ads Help describes Performance Max for travel goals as a way for property advertisers to serve across YouTube, Display, Search, Discover, Gmail, Maps and the Travel channel, using property lists you build in Google Ads or link from a Hotel Center account. If you sell rooms or pitches, that is a genuine fit. For a DMC with a handful of high-value enquiries a month it is riskier: it absorbs brand traffic, reports it back to you as performance, and hides which trip type actually earned. If you run it, exclude brand terms and keep a Search campaign alongside as the control.
Demand Gen creates demand rather than harvesting it, across YouTube, Discover and Gmail. In travel that means shoulder-season pushes, a new region you are establishing, or fixed departures that are not filling. Judge it on assisted enquiries and branded search lift over a window at least as long as your median enquiry-to-deposit time, not on last-click bookings inside the month you spent the money.
Pace budget against the booking curve, not the travel calendar
Most operators I talk to budget against when people travel. Your ad budget belongs on the other axis: when people book. The two curves can sit five months apart, and spending against the wrong one means arriving at the auction after the decision has been made.
Plot two or three years of confirmed bookings by booking date rather than departure date. You will almost certainly find more than one curve inside the same business, and the reasons are structural rather than mysterious. A long-haul tailor-made trip needs international flights, annual leave signed off, sometimes visas and vaccinations, and a deposit big enough that people want thinking time, so it books far ahead. A camping pitch needs none of that, so it books close in and moves with the weather forecast. Fixed departures make a third shape again, because they fill and then stop selling entirely. Read those curves off your own bookings rather than borrowing anyone's rule of thumb.
- Push hardest in the run-up to your booking peak, not during it. Work backwards from the peak by at least your median enquiry-to-deposit time, so the money is in the auction while people are still choosing between operators.
- Never zero out the shoulder months. That is where your cheapest high-intent clicks live, and switching off costs the bidding system its learning.
- Keep a reserve for the late-booking tail, and size it from your own data: take the share of last year's bookings that arrived inside the final few weeks before departure, and hold back roughly that share of the annual budget.
- With fixed departures or occupancy limits, pace against remaining inventory rather than a flat monthly figure. A departure that is 80 percent sold deserves different money to one that is 20 percent sold.
Buy source markets, and land them on a page that matches
Geo targeting deserves five minutes every quarter, because the default causes an expensive travel problem. Google Ads Help lists presence or interest as the recommended location option, which means your Namibia campaign can serve to people who are in Namibia or have merely shown interest in it. If you sell to German and Dutch travellers, that is a steady drip of clicks from inside the destination and from people researching a country they will book through somebody else entirely. Set targeting to presence in your source markets, and exclude the destination unless you genuinely sell on the ground.
Language targeting is the sibling mistake. Google Ads Help says it uses a variety of signals to work out which languages someone understands, including the language of the query, the user's settings and other signals derived by Google AI, rather than the language of your landing page. Dutch and Scandinavian travellers routinely research in English, so a campaign restricted to one language can quietly exclude the people most likely to buy. Then the landing page, where most of the waste happens: if the ad promised a 10-day Namibia self-drive, the page opens with a 10-day Namibia self-drive, not a homepage carousel. Every trip-type page needs:
- The exact trip from the ad above the fold, with a route map or day-by-day outline.
- A price anchor: from X euros per person, what is and is not included, and the deposit.
- Honest lead time and next available departures, so nobody enquires about something you cannot operate.
- A named human with a photograph and real expertise in that region.
- Proof that carries weight in travel: verified reviews, local partner credentials, financial protection.
The mistakes I see most, and what to do instead
- Optimising the whole account to one cost per lead. Instead, value enquiries per product line and move to value-based bidding.
- Leaving the conversion window at its default on a months-long cycle. Instead, extend it and import an earlier stage as your bidding signal.
- Bidding on bare destination names. Instead, buy searches naming a trip shape, duration, party type or operator.
- Pausing everything in the off-season. Instead, hold a maintenance budget so bidding keeps learning and early bookers still find you.
- Judging a campaign after two weeks. Instead, judge over a window as long as your median enquiry-to-deposit time, using enquiry quality as the early read.
Start here, in this order
If you do one thing this week, make it measurement. None of this is fast, which is rather the point: nobody copies a pipeline where every enquiry carries its click identifier and every trip type has a defensible margin behind it.
- Week one: capture the click identifier on every enquiry form, including the brochure download and the callback request, and store it against the CRM record. Nothing else here works without it.
- Week one: pull two or three years of bookings by booking date and trip type, then work out average margin and enquiry-to-deposit close rate for each. That table is your bidding brief.
- Week two: set geo targeting to presence in your source markets, add the negative themes, and check conversion windows against your real lead time.
- Weeks two to four: split the account by trip type, give each its own landing page with the itinerary and price anchor on it, and set targets from the margin table.
- Weeks three to eight: build the daily upload of stage changes with values against the current Data Manager API, then let it report for a full cycle before you touch bid strategy.
- From week eight: move campaigns with enough data onto value-based bidding one at a time, holding each for at least one conversion cycle.
- Every quarter: recalculate stage values from closed business, replot the booking curve by booking date, and reset budgets against it.