Are you buying growth or paying for bookings you'd get anyway? What the data shows

If you run paid media for a travel company, you have likely looked at last year's performance and asked some version of the same question: am I actually buying growth, or am I paying for bookings that would have happened anyway?

It is the hardest question in travel marketing. And looking at how 120-plus brands deployed more than $250 million ad spend, the data suggests most operators may be answering it with the wrong instinct.

For multi-day tour operators specifically, the 2025 Propellic Travel Paid Media Benchmark Report delivered a compelling conclusion: The window of peak booking activity is also the window of worst paid media efficiency. The very moment these operators are scaling spend hardest is the moment that spend is buying the least.

High-booking-season campaigns delivered an average ROAS (return on ad spend) of 2.6x. Shoulder-booking-season campaigns delivered 6.3x. Travelers booking in shoulder season spent within 5% of what they spent at peak, so the operators were not just selling cheaper trips. They were selling roughly the same product to the same kind of customer for less than half the customer acquisition cost.

Picture how that plays out on the ground. In July, 50 multi-day operators are bidding for the same "safari tours Kenya" keyword. A click that cost $3 in February now costs $11. Conversion rates barely move because the traveler clicking in July is no closer to booking today than the traveler who clicked in February. They are still researching, still comparing, still talking to spouses and friends. Multi-day travel has one of the longest decision windows in the industry, often stretching months from first ad click to final booking. Bidding aggressively at peak booking demand pays a premium for clicks that will not actually book any faster than the cheaper clicks placed in slower months.

The pattern is sharpest in multi-day because its booking arc is longest. Online travel agencies show a milder version of the same effect. Day tours follow the more intuitive pattern: For that sector, peak booking demand and peak paid efficiency line up, because day-tour bookings happen fast enough that the auction premium does not decouple from the booking. The average day-tour decision window is just 1.9 days.

The search trap

Another finding from the data is more uncomfortable, because it is a problem most operators are funding themselves and one that requires a real overhaul of the measurement stack to fix.

For those less familiar, non-brand keywords are the broad, exploratory phrases travelers use before they have picked a brand. Searches like "best Galapagos cruises" or "Iceland northern lights tour" rather than "Preferred Hotels" or "G Adventures." These are the terms that bring new customers into the funnel for the first time, which is why they account for the majority of paid media spend across every sector the report covers. In multi-day tours, non-brand makes up 90% of total spend. In day tours, 85%. Even OTAs sit at 83%.

Then look at the conversion data. In multi-day, non-brand keywords account for 90% of paid spend but generate only 74% of paid conversions, a 15-point spend-to-conversion gap. Day tours show a 12-point gap. The implication is not that non-brand investment is wasted, of course. The implication is that travel brands run the risk of paying for the awareness layer of the funnel without building the systems to capture the demand they have stirred up.

Here is what that looks like in practice: A traveler searches "best Galapagos cruises" and clicks a paid ad. They browse three trip options on the operator's site, read some reviews, then close the tab. Any travel marketer has seen this 10,000 times in their analytics. Two weeks later, after watching YouTube videos and asking a friend who has been, that same traveler types the operator's name directly into Google and books through the brand search result.

In most attribution systems, the brand keyword gets credit for the booking. The non-brand ad that initiated the journey looks like it underperformed. Part of the spend-to-conversion gap reflects exactly that attribution reality.

The travel brands pulling ahead of the pack are not spending dramatically more or less on non-brand than the rest of the category. They are just measuring it better. They have built the attribution infrastructure to credit non-brand spend for the bookings it actually drives downstream, and they treat prospecting and remarketing as a single funded system rather than two competing line items fighting for quarterly ROAS optics.

Three sectors, three stories

The biggest shift in the 2025 data is that there is no single industry-wide answer to the question of where top-quartile operators outperform. The size of the gap between average and top-quartile performance (what the report calls P75, the threshold above which the top 25% of contributors sit) varies dramatically by sector. The variance itself is the more useful finding.

In OTAs and marketplaces, the gap between the average performer and the top quartile is 68%. The mean OTA operator runs at 3.36x ROAS; the top-quartile operator clears 5.65x. The contributor spread is wider in OTAs than in any other sector covered, which means the headroom for an average OTA to materially improve performance is substantial.

In day tours and experiences, the gap is 31%. The mean operator runs at 2.92x; the top quartile clears 3.83x. Meaningful, but the distribution is tighter. Day tours is a sector where execution discipline pays off rather than one where a few brands occupy a fundamentally separate performance tier.

In multi-day tours, the gap is just 4%. The mean operator runs at 4.55x; the top quartile at 4.73x. The sector is operationally consolidated. Most contributors are running paid media at broadly similar efficiency, which means the strategic question for a multi-day operator is not "how do I break into the top quartile" but "how do I extract more from the operating model already running."

These sector-level gaps also show meaningful variation when contributors are tiered within each sector by average order value. The implication for any travel marketing leader looking at their own numbers is that "average paid media performance" in travel does not mean one thing. It means three different things depending on the sector, and meaningfully different things again once tier is accounted for within a sector.

Across the sectors where meaningful top-quartile lift is available, the leading operators tend to share a set of operational patterns we see firsthand. They own category-level search terms beyond their own name, capturing demand before competitors do. They scale spend during shoulder season when competitors retreat. And they use attribution models that credit ads which influenced a booking but did not directly close it, rather than the last-click view that quietly defunds upper-funnel investment. The full 2025 benchmark report names six operational patterns in total, with the specific channel and budget mechanics behind each.

The honest answer to whether you're buying growth or paying for bookings you'd get anyway depends on whether your measurement stack can see far enough upstream to tell the difference - and the gap between the brands that can and the brands that cannot is now the defining feature of each category.

The 2025 dataset offers a baseline against which travel marketers can evaluate their own paid media operations in what is shaping up to be a more volatile 2026.