Measuring the impact of marketing in travel has never been straightforward. The purchase is considered, the journey is long, and the moment someone decides to book rarely looks anything like the moment they first encountered your brand.
What we’ve found from the data and conversions we’ve tracked across travel brands is that this gap, between where the journey starts and where it ends up, is wider, and more consequential, than most marketing teams realise.
This piece works through the data we’ve collected to show how travel customers actually behave, where the money is going, what’s converting, what isn’t, and what it all means for the way travel brands think about measurement.
💡 Key insights
Google Paid Search delivers the lowest cost per conversion, but it’s mostly capturing existing demand rather than creating it. It generated an average cost per conversion of £61.25, less than half the industry average of £127.06, but much of this demand has already been influenced by upper-funnel marketing.
Upper-funnel channels appear less efficient than they really are. Facebook, Instagram, TikTok and YouTube report higher CPAs, yet much of their impact occurs weeks before a booking and often goes uncredited in standard attribution models.
Direct traffic is likely overstated. While 25.8% of all sessions are recorded as Direct, much of this traffic is the final step in journeys that began through social, content, display or other marketing channels.
Standard attribution misses a significant share of bookings. Offline sales, long consideration journeys and custom booking systems mean many high-value conversions never get connected back to the marketing activity that influenced them.
How travel customers arrive and what that tells you about attribution
Across the travel and hospitality brands we’ve analysed, total sessions came to just over 10+ million. The split between channels is revealing, and in some ways counterintuitive.
| Channel | Share of traffic sessions % |
| AI Referral | 0.1% |
| Direct | 25.8% |
| 17.2% | |
| Organic Search | 18.0% |
| Paid Search | 21.9% |
| Programmatic | 0.5% |
| Referral | 3.1% |
| Social Organic | 2.3% |
| Social Paid | 11.0% |
This table doesn’t include offline channels, but it’s worth noting that 11.7% of total conversions tracked also came from offline sources such as print and TV, activity that is typically missed by standard digital attribution tools.
Direct traffic accounts for 25.8% of all sessions
Direct traffic takes the largest single channel share in the dataset, and on the surface, a number that looks healthy.
In practice, it’s one of the most misunderstood traffic sources in travel marketing.
Related: What is direct traffic in Google Analytics 4 and how we navigate it
A significant portion of what arrives as direct is the end point of a journey that started somewhere else entirely, a social ad seen weeks earlier, a piece of content that planted the idea, a recommendation from someone they trust.
By the time the customer comes back to book, the channels that influenced the decision have long since dropped out of the picture. For travel brands dealing with consideration windows beyond 90 days, this matters more than in almost any other sector.
Upper-funnel channels, paid social, display, CTV, content, are doing the work of building intent and shaping preference across that entire window, but because the booking happens long after the last touchpoint, they get little or no acknowledgement.Â
What the data shows instead is direct traffic and brand search, which look like the drivers of performance but are really the final step in a much longer story.
Paid search accounts for 21.9% of sessions
Paid search generates the highest conversion rate of any significant channel at 2.24%, and it isn’t surprising. Paid search captures intent that’s already been formed, the customer has already decided they want to go somewhere, they’re now comparing options. It deserves credit for closing. What it rarely deserves is full credit for starting the journey, but that’s often what it gets.
Organic search accounts for 18% of sessions with a conversion rate of 1.89%
This is marginally above the overall average of 1.71%. What’s interesting here is the call volume. From the conversions we’ve tracked, organic search generates a meaningful proportion of phone calls, not just form fills.
This matters because organic is often treated as a top-of-funnel channel and undervalued in conversion reporting. In travel, where customers frequently call to complete complex or high-value bookings, organic search is doing more commercial work than most attribution models give it credit for.
Email sits at 17.2% of sessions with a conversion rate of 0.67%
Email drives a significant volume of traffic but converts at less than a third of the paid search rate. There are a few possible explanations. Email audiences include a mix of existing customers, past enquirers, and cold contacts, some of whom are being retargeted on journeys that started elsewhere. The conversion rate looks low partly because email is often doing nurturing work rather than closing work. Whether that’s being accounted for in budget decisions is another question.
Social paid accounts for 11% of sessions with a conversion rate of 1.78%
These benchmarks are below paid search but above the overall average. Social organic trails significantly at 2.3% of sessions and a conversion rate of 1.18%. The gap between paid and organic social in travel is consistent with what we see broadly. Social organic builds brand presence; social paid drives measurable traffic. Neither tells the full story of what social is actually contributing to bookings, because a significant proportion of social influence happens at the impression level, before a click is ever made.
AI referral accounts for 0.1% of sessions but converts at 2.86%
While it accounts for the smallest share of traffic in the dataset, AI referral has the highest conversion rate of any channel we’ve tracked. Customers arriving via AI-generated recommendations or answers are arriving with a high degree of intent and a high degree of trust in what they’ve been told. As AI search continues to develop, this channel will grow. Whether travel brands are set up to track and attribute it properly is a question worth asking now, before the volume gets significant enough to matter materially.
The phone call problem and why it’s bigger than most teams think
From the conversions we’ve analysed across travel, 28.4% of all conversions come in over the phone. Online bookings account for 71.6%. That’s roughly 1 in 3 bookings arriving via a channel that most digital attribution models either miss entirely or handle poorly.
| Industry | Share of forms % | Share of calls % |
| Travel | 71.6% | 28.4% |
The calls that come in represent real bookings, often high-value ones, long-haul trips, group travel, and complex itineraries. The kind of bookings where a customer wants to speak to a person before committing several thousand pounds.
What makes this particularly acute in travel is the nature of the booking engine landscape. Many travel brands operate on custom-built reservation systems or proprietary booking engines that don’t connect natively to their marketing stack.
The system knows the booking happened, but the marketing team often has no idea which campaign, channel, or touchpoint drove it, and the result is that a significant portion of the most commercially valuable bookings, the ones with the highest average order value, are effectively invisible to any attribution model trying to make sense of performance.
Related: How to integrate offline data into your digital strategy & targeting
Most teams we speak to are aware of the gap, the harder question is what to do about it. First-party tracking that captures the customer journey from first visit through to call completion, matched back to the original marketing source, can close this gap.
The technical challenge is real, particularly for brands running custom systems. But the commercial cost of not solving it, optimising away from channels that are driving high-value calls because they look unproductive in the data, is significant.
The online-to-offline gap
Phone calls are only part of the story. A meaningful share of travel bookings are researched online and closed offline entirely, in a branch, with an agent over the counter, at a kiosk, or through a sale that gets keyed into a point-of-sale or CRM system with no digital record of the journey that led up to it.
This is a different problem to the phone call one. A call, even an untracked one, at least leaves a record somewhere in the phone system. A customer who spends three weeks comparing destinations on their phone, clicks through from an ad, opens two remarketing emails, and then walks into a shop to book, leaves no evidence.
As far as any digital attribution tool is concerned, that journey never happened. The booking shows up in a completely separate system, if it’s captured as a “conversion” at all.
Closing this gap generally means connecting the online journey to whatever system offline bookings land in, whether that’s a retail POS system, an agent’s booking tool, or a CRM. Without that link, any assessment of channel performance is built only on the conversions that happen to be visible, not the ones closing offline
💡 Pro Tip
One travel brand combined first-party journey tracking with offline conversion data and marketing Mix Modelling, to connect digital interactions to offline bookings, measured the combined impact of online and offline channels, and identified diminishing returns across its media mix. The result was more than $400 million in attributed revenue, a unified ROAS of 8x across all channels, and a 100-week historical model that gave the team confidence to optimise future investment.
How a travel brand unified MMM and MTA to connect online and offline
How travel brands are allocation ad spend
Total ad spend across the travel brands we’ve looked at runs to an average of £231,566.71 annually, with a median monthly spend of £7,337.74. The average cost per conversion across the industry sits at £127.06, though as we’ll come to, what counts as a conversion in this figure matters a great deal.
| Channel | Average total ad spend by channel | Average CPC by channel | Average CPL by channel |
| Google Paid | £945,560.31 | £1.85 | £61.25 |
| Bing Paid | £31,337.84 | £1.66 | £45.97 |
| Facebook Paid | £290,930.76 | £1.23 | £161.84 |
| Instagram Paid | £84,367.51 | £3.45 | £102.53 |
| TikTok Paid | £17,504.94 | £2.37 | £169.68 |
| YouTube Paid | £19,698.82 | £4.55 | £221.10 |
| Total | £1,389,400.18 | £2.25 (avg. CPC) | £127.06 (avg. CPL) |
Google paid search dominates spend, accounting for the largest share at £945,560.31 in total yearly spend across the dataset. The average CPC of £1.85 is the lowest of any paid channel, and the average cost per conversion of £61.25 is less than half the industry average. On those numbers alone, Google paid looks like the obvious place to put money.
But there’s a catch, and it’s a familiar one. Google paid search is good at capturing demand that already exists, but it rarely creates it. The customers clicking paid search ads have already decided they want a holiday, they’re now in the comparison phase. The channels that put the idea in their heads, that made them think about a particular destination or brand, don’t show up in Google’s attribution. They’re upstream, invisible, getting no recognition.
Facebook paid sits at £290,930.76 in total spend with a cost per conversion of £161.84, more than twice Google’s figure. Looked at purely through a last-click or even data-driven lens, Facebook looks expensive.
However, Facebook is doing something different. It’s reaching people who weren’t actively looking, building awareness, surfacing inspiration. The cost per click on Facebook (£1.23, the lowest in the dataset) reflects the fact that it’s buying reach, not intent. The conversion cost is higher because the customer isn’t ready to convert yet, they’re at the beginning of a journey that might take two months to complete.
Instagram (£3.45 CPC) and TikTok (£2.37 CPC, £169.68 cost per conversion) show similar dynamics. These are channels where the influence is real but the attribution is partial. From the data we’ve looked at, social paid channels generate a conversion rate of 1.78%, but that’s almost certainly understating their contribution to bookings that ultimately complete via direct, brand search, or phone.
YouTube shows the most interesting anomaly in the dataset. At £4.55 CPC, the highest of any channel, and a cost per conversion of £221.10, it looks expensive by any standard measure.
But YouTube in travel is functioning just like television, building brand, driving consideration, creating the emotional pull that makes someone open a new tab and start searching. Measuring it on last-click or even short-window attribution produces a number that is almost certainly wrong, and in the wrong direction.
The pattern across the spend data is consistent. The channels that create demand are being measured by standards designed for channels that capture it. The result is predictable, budgets migrate toward the bottom of the funnel, demand creation weakens, and at some point the pipeline starts to thin.
What sits behind the conversion rate numbers
The overall conversion rate across the dataset is 1.71%. Across channels, the range runs from 0.67% (email) to 2.86% (AI referral), with paid search at 2.24% and organic search at 1.89%.
| Channel | Conv rate |
| AI Referral | 2.86% |
| Direct | 1.87% |
| 0.67% | |
| Organic Search | 1.89% |
| Paid Search | 2.24% |
| Programmatic | 1.20% |
| Referral | 1.50% |
| Social Organic | 1.18% |
| Social Paid | 1.78% |
| Total | 1.71% |
A few things are worth noting here. First, these conversion rates include both form submissions and phone calls. That matters because phone calls are disproportionately high-value.
A conversion rate that treats a callback request and a £6,000 long-haul booking enquiry as equivalent events is technically accurate but commercially misleading.
Second, the conversion rate of 1.71% is an average across a wide range of journey stages and intent levels.
Customers arriving via paid search are much further along the funnel than customers arriving via social paid. Comparing their conversion rates without that context produces misleading conclusions about channel efficiency.
Third, some of these numbers say more about scale and measurability than about performance. Programmatic sits toward the bottom of the table at 1.20%, and it’s tempting to read that as a channel that simply doesn’t convert well.
But it’s also the smallest channel in the dataset by traffic share, just 0.5% of sessions, so there’s barely any of it happening in the first place. A display ad seen while reading the news, or a CTV ad seen mid-programme, rarely gets clicked; its influence tends to show up later, in a branded search or a direct visit, by which point any link back to the original ad is gone.
Compared with a channel like Facebook and YouTube, which at least leaves a trail of clicks and views, programmatic often leaves almost nothing for a last-click model to credit. The low conversion rate here reflects that measurement gap as much as it reflects genuine underperformance.
Related: How to measure top of the funnel marketing and brand awareness
Lastly, and probably most importantly, these are the conversions we can see. The phone calls that came in from customers who had been through a 90-day journey involving five different channels, two devices, a visit to a travel agent, and a conversation with a friend, those bookings are in the dataset somewhere. But which marketing activity drove them is, in most setups, unknown.
Traffic volumes in travel and what the median actually shows
| Industry | Median monthly traffic |
| Travel & Hospitality | 3,512 |
The median monthly traffic figure for travel sits at 3,512 sessions. Set against the total sessions in our dataset, this reflects significant variation in scale across the brands we’ve analysed.
Some travel brands are operating at considerably higher volumes; others are much smaller. The median is a useful reality check against averages that can be skewed by larger operators.
What it also reflects is that for a significant number of travel brands, the total addressable audience arriving at their digital properties each month is smaller than it might appear.
At a 1.71% conversion rate and a median of 3,512 sessions, that’s around 60 conversions per month. At an average cost per conversion of £127.06, that’s roughly £7,600 in acquisition cost, broadly consistent with the median monthly ad spend of £7,337.74.
The maths is tighter than it looks. And when 28% of those conversions are coming in via phone, potentially untracked, potentially unattributed, the efficiency of that spend is harder to assess than the headline numbers suggest.
What this means for how travel brands should think about measurement
The data, taken together, points toward a few consistent conclusions.
Upper-funnel channels are contributing more than they appear to. Paid social, display, YouTube, and AI referral are all doing work that doesn’t show up cleanly in standard conversion reporting. The instinct to cut these channels when CPA looks high is understandable, but based on incomplete information.
Phone calls are a part of the conversion picture. At 28.4% of all conversions, and disproportionately weighted toward high-value bookings, call tracking isn’t optional for travel brands. It’s foundational, and for brands running custom booking systems or proprietary CRMs, connecting those calls back to marketing source requires technical investment.
Offline conversions are a blind spot most brands haven’t measured at all. Where phone calls at least generate a record somewhere, bookings that are researched online and completed in-branch or with an agent often leave no digital record whatsoever in traditional attribution models.
Direct traffic is inflated. A quarter of all sessions arriving as direct is a measurement problem as much as a channel mix story. Understanding what’s actually driving direct, which campaigns, which channels, which touchpoints, requires a first-party tracking approach that follows the customer across the full journey, not just the final session.
The consideration window is long enough that short attribution windows are structurally inadequate. A 7-day click window, or even a 30-day window, misses most of the travel booking journey. The channels active in weeks two through eight of that journey, the ones shaping the decision, leave no footprint in standard attribution, and the gap that creates isn’t something better data collection alone can fix. It requires a different measurement approach entirely, one that combines click-based attribution with impression modelling and, where possible, marketing mix modelling to surface the channels that standard tools can’t see.
None of this means the data is useless. It means the data needs context, and a measurement framework that’s built around how travel customers actually behave, rather than how digital attribution models assume they do.
Ruler can help make sense of it all
The measurement challenges in travel aren’t new, but they’re getting harder to ignore as budgets tighten and the pressure to evidence performance increases. At Ruler, we work with travel brands to close the gaps this piece has outlined, connecting every conversion type back to its marketing source, whether that happens online, over the phone, or through a booking engine that was built in-house.
Book a demo and we’ll show you what your performance picture could look like when measurement is unified and the gaps are closed.
