The Traffic Paradox: How Tour Operators Are Booking More Trips with Fewer Website Visitors

Your Traffic Is Down. Your Bookings Might Not Be.

If you’ve opened your website analytics recently, you’ve probably noticed something unsettling: organic search traffic is falling. Across the 275+ tour and activity operators we work with at TOMIS, total website sessions dropped roughly 13% year over year in 2025 compared to 2024.

That’s a real number, and it represents real people who aren’t landing on your site. It makes sense to worry about it.

Here’s what makes less intuitive sense: over that same period, conversion rate — the share of visitors who actually book — climbed 6.5%. The visitor who reaches a tour operator’s website today is meaningfully more likely to buy than the one who visited a year ago.

This isn’t a coincidence. It’s the direct result of how AI has quietly reorganized the top of the travel planning funnel.

What Changed (and Why)

A few years ago, your website served two audiences simultaneously: people doing early research (“what should I pack for a rafting trip?”) and people ready to book (“book Colorado river rafting this Saturday”). Both groups arrived through organic search, and you couldn’t easily tell them apart. You optimized for traffic, because traffic meant opportunity.

AI assistants changed that equation. ChatGPT, Google’s AI Overviews, Perplexity, and similar tools now absorb the majority of those early-stage, informational queries before a traveler ever visits a website. The person who used to Google “what to wear on a river trip” and click through to five different pages now gets that answer from an AI in thirty seconds and never loads your FAQ.

What’s left in your analytics is a more selective group: people who’ve already done their research, narrowed their options, and are now looking for a specific operator to book with. They’re not browsing — they’re deciding.

That’s why the traffic line fell while the conversion line rose. The browsers filtered themselves out upstream.

The Numbers, Channel by Channel

The GA4 data across our client portfolio tells the story clearly when you break it down by traffic source.

Organic Search saw the steepest session decline — down about 26% year over year. Transactions fell roughly 17% alongside it, and organic revenue was off about 15%. The gap between the session drop and the transaction drop matters: organic traffic lost volume faster than it lost buyers. The people still clicking through from search results are better prospects than the average organic visitor was a year ago.

Paid Search sessions were down about 7% — but transactions were up 4%, and paid search revenue grew nearly 14%. Fewer clicks, more bookings, meaningfully higher revenue per click. If you’re running search campaigns and grading them by click volume, you’re missing this story entirely.

Cross-Network (Performance Max) is where the growth concentrated. Sessions grew about 38% year over year, and transactions more than doubled — up 133%. Google’s automated, multi-surface campaign format found buyers across Search, Maps, YouTube, and Display in a way that standalone search campaigns couldn’t match. For tour operators with active PMax campaigns, this was the year it stopped being an experiment.

Paid Social sessions grew 40% and transactions grew 116%. Paid social has become the new top of the funnel — the channel where travelers first encounter an operator after an AI has already done the early filtering. The intent arriving at a paid social ad in 2025 is different from what it was in 2023.

Direct Traffic — the bucket that catches a lot of AI-referred visitors, since AI tools typically don’t pass clean referral data — was down about 15% in sessions and transactions. But within that bucket, an increasing share of visitors are coming from AI-generated recommendations that show up as “direct.” We track this separately.

AI Referrals: The Fastest-Growing Traffic Source You’re Probably Not Measuring

Across the TOMIS portfolio, dedicated AI-referred sessions — visits attributable to ChatGPT, Perplexity, Google’s AI answers, and related tools — grew from about 6,300 sessions in full-year 2024 to more than 121,600 in 2025. That’s roughly a 19x increase in a single year.

The growth wasn’t a spike. It climbed every month from January 2025 onward, peaking in summer and sustaining a new, higher baseline into fall and winter. In January 2024, AI referrals were a rounding error — 150 sessions across the entire portfolio. By July 2025, the same metric crossed 15,000 sessions in a single month.

It’s still a small share of total traffic. But it’s growing faster than any other channel we track, and unlike most traffic channels, it’s been moving in a straight line upward rather than fluctuating seasonally. The operators capturing AI referral traffic today are building a baseline that compounds.

Why AI Engines Cite Some Operators and Not Others

AI tools don’t recommend operators randomly. They surface sources that are easiest to parse, trust, and cite — pages that answer real traveler questions directly, are structured in a way AI can interpret, and look authoritative on the specific things people ask about.

In practice, that means:

• Clear answers to common questions — what to bring, age and fitness requirements, cancellation and weather policies, what happens if water levels are too high
• Schema markup and structured data — telling search engines and AI tools exactly what your trips cost, when they run, and how to describe them
• Specific product pages — detailed, accurate descriptions of individual experiences rather than thin, generic category copy

The operators seeing the strongest AI referral growth built these foundations over the past two years, often as part of broader SEO work. The result is showing up now in AI recommendations, and those AI recommendations are sending visitors who are unusually close to booking.

The Catch: You Can’t Schedule AI Visibility

Being cited by AI is valuable, but it’s not a dial you can turn. You can create the conditions for it — better content, better structure, better authority signals — but you can’t buy a slot in ChatGPT’s recommendations the way you can buy a paid search position.

That’s the strategic gap. Paid search and Performance Max give you a lever you can actually pull when you need bookings next Tuesday. AI visibility gives you organic demand you can’t time. The operators winning right now pair both: building for AI citability as a long-term foundation while running paid channels as the mechanism they actually control.

The Right Scoreboard

If organic traffic is no longer a reliable proxy for business health, what should tour operators be measuring?

Conversion Rate on Organic Traffic

Rising conversion rate on falling sessions is the healthiest signal in this environment. It means you kept the buyers when the browsers left. If your organic conversion rate is climbing even as session volume drops, your site and content are doing what they should.

Branded Search Volume

When an AI tool mentions your company by name and a traveler then searches for you directly, that shows up in branded search. It’s demand that AI created and you captured. Watch your branded search trend in Google Search Console — it’s one of the clearest leading indicators of AI-driven awareness.

Bookings Per Session

This metric — sometimes called conversion rate, but tracked over time and across channels — tells you whether your marketing is becoming more efficient. Portfolio-wide, bookings per session has been trending up even as raw traffic declined. That’s the number that reflects what’s actually happening to the business.

Revenue Per Visitor

Similar logic. If each visitor is more likely to book, revenue per visitor should be climbing. Track it channel by channel — paid search and cross-network typically run significantly higher than organic or social, and the gap has been widening.

Where the Budget Shift Is Happening

Here’s the part most operators are slow to act on: once you measure bookings per session rather than raw traffic, paid channels look completely different.

A campaign that “only” sends 800 clicks but closes at 8% is outperforming a blog post that drives 8,000 sessions at a 0.5% booking rate. The right scoreboard makes the case for paid almost on its own.

The operators who’ve captured this shift are doing three things with their paid budgets:

1. Protecting branded and high-intent search terms. When a pre-qualified traveler finally types “book [your activity] near [your location],” they’re at the moment of decision. Paid search puts you at the top of that moment. Relying on organic rank alone — especially as AI Overviews push organic results further down the page — means leaving high-intent clicks to competitors.

2. Leaning into Performance Max. The data across our portfolio shows PMax transactions more than doubling while paid search transactions were essentially flat. Google’s multi-surface automation is finding buyers in channels and at moments that keyword-based search campaigns miss. For tour operators with solid conversion tracking and a meaningful catalog of trips, PMax has become the highest-leverage campaign type.

3. Treating paid social as a top-of-funnel channel, not a direct-response channel. Paid social transactions more than doubled in our portfolio last year, but the path to conversion is different from search. Social intercepts a traveler who’s been pre-qualified by AI earlier in their research and surfaces the operator at the awareness stage. The booking often happens later — through direct, branded search, or a return visit. Measuring paid social purely on last-click purchases misses most of its value.

What to Do With This

None of this requires panic, and none of it requires starting from scratch. Most of it is a question of what you measure and where you point your attention.

If your organic traffic is down but your conversion rate is rising, that’s a healthy business in a changing environment — not a crisis. The signal to be concerned about is organic traffic falling and conversion rate falling at the same time. That means you’re losing visitors and the ones who remain aren’t buying, which points to site experience or offer problems independent of the traffic environment.

If your paid channels are flat year over year, you’re very likely leaving efficiency on the table. The same budget is landing on a more qualified audience than it was two years ago. The question isn’t whether to spend more — it’s whether your campaign structure, bidding strategy, and conversion tracking are set up to capture the efficiency that’s already there.

If you haven’t started tracking AI referrals separately, the path is straightforward: segment your “direct” channel by source in GA4 and look for traffic from claude.ai, chatgpt.com, perplexity.ai, and similar domains. Set it up as a custom channel grouping and watch it monthly. What’s a rounding error today will be a meaningful number by next summer.

The One-Paragraph Version

Website traffic is becoming a vanity metric in tourism. AI assistants now do the early research and filtering that used to happen on your website, so fewer visitors arrive — but the ones who do are far closer to booking. Across the 275+ operators we work with, that shows up as total sessions down roughly 13%, while conversion rates are up 6.5%, paid search revenue is up 14%, and cross-network (Performance Max) transactions more than doubled year over year. AI-referred visits grew 19x in a single year. The channels where targeting is precise — paid search, Performance Max, paid social — are converting buyers far more efficiently than they did two years ago, while AI citability has become a core content deliverable rather than a bonus. The operators winning right now stopped grading themselves on traffic and started grading on bookings per visitor — and moved their budget toward the channels that produce them.