Listen to the podcast episode
Or listen now on Apple Podcasts or Spotify.
Key Takeaways
Surface-level analytics tells you almost nothing useful: Most managers open Google Analytics, glance at total sessions and total purchases, check the box and move on. That tells you whether the number went up. It doesn’t tell you where in the booking process people are falling out, which is the only thing you can actually fix.
Measure every step of the booking funnel, not just the ends: The whole point of a funnel view is watching the fall-off between steps. When you look at each stage separately, a channel that looks dead on revenue often turns out to be pushing people two or three steps deeper than you assumed.
- Session start / homepage view
- Date search entered
- Search results page viewed
- Property detail page viewed
- Checkout started
- Purchase completed
Segment the funnel by channel before you judge any channel: Conrad’s example: 200 people click through from a newsletter, 110 run a date search, 47 open a property page, 2 start checkout. Zero bookings that session, so it reads as a failure. It isn’t. You just moved 200 people from booking-unaware to deep in the funnel in one click.
Email and Meta are top-funnel channels graded on a bottom-funnel scorecard: A newsletter to past guests is an awareness play, not a closing play. The exception is checkout abandonment email, which is genuinely bottom-funnel. Meta is the same story: a five-second video shows a guest more about your property than 200 Google searches would, even when it never gets last-click credit.
Slice by device, platform, and traffic source and most problems surface: Run your funnel three ways over a two-to-three month window and compare: mobile vs. desktop, Android vs. iOS, then Google Ads vs. organic vs. Meta. It sounds too simple to work. It isn’t. Those groups behave completely differently, and the gaps show you where to look.
People landing deep in your site need context or they leak: Roughly half your traffic starts on the homepage. The rest drops straight onto a property or category page with no destination context, and if that page doesn’t show them you have 100 other cabins, they assume the one listing is all you have and bounce. Your own settings often cause it: Performance Max final URL expansion sends loosely matched queries to whatever page Google picks, and indexing does the same when a single listing outranks your destination page for a high-level term.
Resort-page drop-off is usually a layout problem, not a traffic problem: Nobody builds a bad page on purpose. The SEO team wants the paragraph, someone wants the property video, someone wants the photo gallery, and the highest-paid person in the room settles it. Every addition pushes the actual units further down until visitors never reach what they came for.
Session recordings find what the numbers only hint at: Watching Clarity sessions at 2x or 4x is tedious and your eyes go buggy, but it’s how you learn why a step is leaking. Even without full sessions, scroll depth and heat maps are worth it, because your site renders differently across mobile devices in ways a report will never show you.
AI traffic sits at the top of your funnel, not the bottom: The industry throws around 40 to 50 percent numbers while actual client analytics show barely any AI-referred sessions. The reason: people use AI before they’ve picked a destination, with prompts no Google search could have handled. For a considered family trip, nobody is handing the final booking to a bot.
Delegate the work, don’t abdicate it: Delegation is someone executing your vision. Abdication is handing it over and hoping. Plenty of vendors in this space are excellent at selling, so trust but verify: go look at your own funnel data, ask hard questions about what’s working, and stay the CEO of your own domain.
What We Cover In This Episode
Conrad and Paul revisit the Google Analytics booking funnel through a 2026 lens: what’s actually changed, why segmenting by channel and device beats staring at top-line sessions, where deep landing pages and resort-level content leak bookings, and where AI-referred traffic really enters the customer journey versus where the industry claims it does.
