Over the past couple of months, we have warned hoteliers about the shifting tides of search—the plateauing of traditional organic Click-Through Rates (CTR) and the aggressive push towards Generative Engine Optimisation (GEO).

But as we sit down with General Managers and DOSM’s to look at their performance dashboards, a new, massive point of friction has emerged. GMs are looking at their Google Analytics 4 (GA4) dashboards, seeing a massive spike in “Direct” traffic and flatlining “Organic Search” traffic, and concluding: “Our brand power is up, so we don’t need to worry about AI search.” They are entirely wrong. Your dashboard is lying to you.

According to our latest YTD 2026 search intelligence dataset, the traditional click-through curve has permanently bent. The old baseline where a position #1 ranking guaranteed nearly a 40% CTR has collapsed to just 27.6% globally, and drops off sharply to 19.2% the second an AI Overview appears on the screen.

The reality of 2026 is that over 70% of the traffic being driven to hotel websites by AI engines is completely invisible in GA4. It is being misclassified, buried, and ignored.

When a high-intent traveller uses ChatGPT, Perplexity, or Gemini to plan an itinerary (e.g., “Find me a 5-star boutique resort in Noosa with private plunge pools and an on-site day spa”), the AI synthesises an answer and provides citation links. When the user clicks those links, they land on your brand’s website or custom microsite. However, because of how mobile AI apps and secure browsers handle data, 70.6% of those sessions have their referrer headers stripped. Without that header, GA4 has no idea where the user came from and it panics and dumps that visitor into your “Direct” traffic bucket.

This is the “Dark AI” Gap… You are winning high-value, commission-free bookings directly from generative AI answers, but your analytics are giving 100% of the credit to a phantom “Direct” browser session.

Marketers who accept their GA4 dashboards at face value are systematically undercounting the commercial value of AI search. According to data from the start of this year, traffic originating from AI platforms converts at an astonishing 1.66% compared to just 0.15% for traditional organic search. Why? Because travellers using AI assistants are no longer browsing top-of-funnel keyword lists. They have a budget, a specific use case, and an intent to buy. They are closer to a booking or an RFQ than any standard Google user.

But right now, your Revenue Manager is likely starving your AI optimisation strategy because they cannot see the ROI on the screen.

To fix your attribution and stop letting the OTAs win this data war, you need to force GA4 to see the machine.

Following Google’s official May 2026 updates regarding default channel definitions, GA4 is beginning to natively flag certain automated and AI referral structures, but it still heavily categorises organic AI clicks as standard “Referrals” or buries them in “Direct” if the header is missing. Here is the exact, manual workaround you should be deploying right now to cleanly segment this data:

Step 1: Build a Custom Channel Group

Do not edit Google’s default channel group. Instead, navigate to Admin → Data Display → Channel Groups and click Create New Channel Group. Name it something specific, like “Default + AI Traffic (2026)”.

Step 2: Write the AI Regex Filter

Click Add New Channel and name it “AI Search”. Set the rule condition to Source matches regex, and paste the following pattern to catch the major players currently driving referral traffic:

chatgpt\.com|chat\.openai\.com|claude\.ai|gemini\.google\.com|perplexity\.ai|copilot\.microsoft\.com|deepseek\.com|meta\.ai|grok\.com

Step 3: Drag it Above “Referral”

This is the critical step most operations teams miss. GA4 evaluates channels sequentially from top to bottom. You must physically drag your new AI Search channel above the standard “Referral” channel. If you leave it below, GA4 will swallow those recognisable AI clicks as a basic referral before your new rule ever evaluates them.

Step 4: Add Landing Pages as a Secondary Dimension

Once saved, view your Traffic Acquisition report and switch to your new channel group. Add Landing Page as a secondary dimension.

This will show you exactly which URLs the AI engines are pulling into their answers. For full-service properties, you will almost always find that the AI isn’t pulling your brand.com home page; it is citing specific, entity-rich MICE hubs, Wedding galleries, or Resort microsites that answer the traveller’s hyper-specific prompt.

Fixing your analytics is step one, but it only tracks the traffic you are already getting. Step two is optimising your digital ecosystem so that you actively win the “Citation” battle over the OTAs.

If your website isn’t structured with advanced schema, direct conversational Q&As, and clear entity signals, ChatGPT and Gemini will simply crawl an OTA’s data library instead and hand them the direct booking.

Stop guessing where your direct revenue is coming from… Click here to request a Free AI Visibility and Dark Traffic Audit for your property.

OmniHyper: Strategic Intelligence for the Modern Hotelier.

With 30+ years across Wyndham, Hilton, Google, and Microsoft, Colin blends hospitality expertise with tech innovation to drive exceptional results.