How to get travel agency clients: five channels ranked by cost per booking

Start with cost per booking, not lead volume

Most travel businesses measure the wrong number. Leads are cheap to generate and easy to feel good about. The number that decides whether a channel deserves your budget next season is the fully loaded cost of one confirmed booking: money spent, plus the hours your team burned handling everything that did not convert.

Include staff time honestly. If a salesperson spends four hours quoting a group enquiry that never books, that is roughly €80–160 of real cost, and it belongs in the channel that produced the enquiry. Companies that skip this step usually conclude that trade shows are cheap and marketplaces are expensive, which is almost always backwards.

The ranking below is deliberately rough. Cost per booking varies by destination, average trip value and how well you sell. What is stable is the order, and the order is what you plan around.

Channel one: repeat clients and referrals

This is the cheapest business you will ever write, and most operators under-invest in it because it does not feel like marketing. A repeat client costs the price of an email. A referred client arrives pre-sold, negotiates less and complains less.

The mechanism is unglamorous. Keep a list of every past traveller with the month they travelled, the party composition and what they spent. Contact them once a year, at the point in the calendar when they historically start planning — not in January because your newsletter tool suggested it. For a summer Mediterranean product, that is usually late autumn to February.

Referrals need a request. Ask at the moment of maximum goodwill, which is two to three days after the client gets home, not at the airport and not six months later. Cost per booking here is functionally zero to €50.

Channel two: trade partnerships with foreign operators and agencies

For a DMC or a ground operator, B2B is usually the backbone. One European tour operator that sends you thirty groups a season is worth more than any campaign you will ever run, and the relationship compounds because switching a reliable ground partner is genuinely painful for them.

The cost is business-development time, not media spend. Expect six to eighteen months from first contact to first booking with a mid-sized operator, and expect the first trip to be a test — small, watched closely, priced tightly. Amortised over a multi-year relationship, cost per booking typically lands between €100 and €400.

The failure mode is concentration. If one partner is more than 30 per cent of your volume, they know it, and your margin reflects that at every renegotiation. Build the second and third partner while the first is still happy.

Channel three: organic search and a website that answers questions

Organic search is the channel with the worst short-term economics and the best long-term ones. Nothing happens for months. Then a page you wrote in March starts producing two enquiries a week in the following February, and it keeps doing that with no additional spend.

What ranks is specific and useful: how a route actually works, what a week costs at three service levels, when to go and when not to, what the permits and licences require. What does not rank is a services page listing every destination you can technically arrange.

Budget realistically. Doing this properly costs either €500–2,000 a month with an agency, or roughly one focused day a week of your own time for a year. Amortised, cost per booking often drops to €50–200 — but only after the payback period, which is typically 12–18 months.

Channel four: marketplaces and lead platforms

A marketplace buys demand at scale and routes it to companies that can service it. You pay for access rather than for the audience-building, which is why the money starts working in weeks instead of quarters. On European Travel Companies, a traveller sends one brief and verified companies reply with priced offers; the contract that follows is between the traveller and the company, never with the platform.

Cost per booking typically lands between €80 and €300 depending on your close rate, which is the part you control. Two companies on the same platform, seeing the same briefs, routinely differ by a factor of three in cost per booking purely because one replies in two hours with a price and the other replies in two days with a brochure.

The honest limitation is that the demand is not yours. It is incremental business you would not otherwise have reached, and it stops when you stop paying. That is a reasonable trade for filling shoulder season; it is a bad foundation for an entire company.

Channel five: paid search and paid social

Paid search works when someone is already looking for exactly what you sell. Terms with clear commercial intent — a named destination plus a service, a named route, a named event — convert. Broad inspiration terms burn money.

European travel keywords with high intent commonly cost €1.50–6.00 per click. At a 2 per cent enquiry rate and a 20 per cent close rate, you are paying for roughly 250 clicks per booking, which puts cost per booking somewhere between €300 and €1,500. That is fine on a €12,000 private trip and ruinous on a €400 day tour.

Paid social behaves differently: cheaper clicks, colder audiences, much longer consideration. It is a reasonable tool for filling a fixed-departure product with clear dates and a clear price, and a poor tool for selling bespoke itineraries to strangers.

Why a thin website destroys paid traffic

The most common way to waste a media budget is to send bought traffic to a site that answers nothing. A visitor who clicks a paid ad arrives sceptical. If the landing page shows three stock photographs, a sentence about passion for travel and a contact form, they leave and you have paid for that.

The pages that convert bought traffic contain the things you would rather not publish: an indicative price band, what is and is not included, who operates the ground services, the licence number, a named human, and photographs from trips you actually ran.

Fix the site before increasing the budget. Doubling spend against a page converting at 0.4 per cent simply doubles the loss.

Trade shows: ITB Berlin, WTM London, IMEX — do they still pay back

They can, but not the way most companies use them. A stand at ITB Berlin or WTM London runs €8,000–25,000 once you include space, build, freight, flights, hotels at conference rates and three or four staff off desk for a week. If that produces four bookings in the following year, your cost per booking is above €2,000.

The companies that make trade shows pay back treat them as relationship maintenance for the trade channel, not as lead generation. They arrive with a diary of twenty-five pre-booked meetings arranged six weeks in advance, most of them with existing or lapsed partners, and they measure the show over three years rather than three months.

IMEX is a different calculation again, because MICE and incentive bookings carry enough value per contract to absorb a high acquisition cost. If your average booking is under €5,000, walk the show as a visitor for two years before you take a stand.

Creators and PR: slow, unpredictable, occasionally decisive

Hosting a creator is a genuine cost — the trip, the staff time, the opportunity cost of the room — and the outcome is not controllable. Most hosted trips produce brand impressions and no measurable bookings. Some produce a year of demand.

The version that works is narrow. A creator whose audience matches your actual product, an agreement in writing covering deliverables and usage rights, and content you are permitted to reuse in your own paid campaigns for at least twelve months. That reuse right is frequently the largest part of the value.

Traditional PR follows the same shape: unreliable individually, useful in aggregate, and worth doing only when a journalist is already writing about your destination. Treat both as portfolio bets, funded from surplus, never as the plan.

Response speed beats every other variable

Across every channel above, the single strongest predictor of whether an enquiry becomes a booking is how quickly a human answers it with something specific. This holds for referrals, for trade, for search and for marketplaces.

The reason is competitive rather than psychological. A traveller planning a trip contacts several companies in one sitting. The first substantive reply sets the terms of the comparison; everyone afterwards is measured against it, and by the third or fourth quote the buyer is tired and defaults to whoever already made things easy.

If you improve one thing this season, make it the path from enquiry received to first priced reply. Halving that time reliably raises conversion more than any budget reallocation you can make.

Seasonality, booking windows and when to spend

Spend follows booking windows, not travel dates. Most European summer leisure travel is researched between January and April, with a second wave six to ten weeks out. Ski is booked September to November. MICE and incentive programmes are decided six to twelve months ahead. Weddings and milestone trips run twelve to eighteen months.

This means a paid budget spent in June for August travel is competing in the most expensive, least productive part of the year. The same money spent in February usually buys two or three times the result.

Organic works the other way. Publish in the quiet months so the pages are indexed and settled before the research season starts. Content published in March for the same March demand arrives roughly a year late.

Which channel to fix first

Work in this order, and do not skip. First, response time and the quality of the first reply — it costs nothing and improves every other channel simultaneously. Second, the past-client list, because it is the cheapest demand available and most companies have never systematically contacted it.

Third, whichever of trade, organic or marketplace suits your product: trade if you sell ground services B2B, organic if you sell a defined and searchable niche, a marketplace if you need enquiries this quarter rather than next year. Fourth, and only once the site converts, paid media.

Two practical constraints. Do not start a new channel in your peak operational month, because it will be under-served and you will wrongly conclude it does not work. And give any channel except paid media at least two full booking cycles before judging it.

• Fix reply speed and reply quality first — zero cost, affects everything.
• Reactivate past clients before buying new ones.
• Choose one growth channel that matches your product and commit for two booking cycles.
• Turn on paid media last, after the website answers real questions.
• Track cost per confirmed booking including staff hours, not cost per lead.