By Pablo Revale
July 16, 2026 · 3 Minutes
Adapted from Originally published in Unstoppable Hospitality on LinkedIn.
A hotel marketer told me paid search was their best channel. But something wasn’t right. Their booking engine lived on a separate domain. The session broke on the way to checkout, and every direct booking got quietly re-labeled on its way through. Paid search wasn’t winning. It was just the last thing GA4 could still see before the lights went out. It was an attribution problem. That story is fiction. Or is it? Only 23% of the hotels that use GA4 have their analytics set up properly. The other 77% are making million-dollar decisions on numbers that don’t describe reality.
Bad tracking doesn’t stay the same size. It grows. A bad month of data is a bad month. A year of bad data is a year of budgets, campaigns, hiring, and bonuses, all stacked on a measurement that was never real. And the longer it runs, the more expensive it is to admit, because admitting it means questioning every decision you made on top of it. So most teams don’t. They keep optimizing toward a number that was never true, and they call the number "performance." Gartner puts the average cost of poor data quality at 12.9 million dollars a year per organization. Most of that isn’t the wrong number on the screen. It’s the stack of decisions made on top of it. You can’t cover the sun with one hand and pretend it’s night.
This is rarely one dramatic failure. It’s small, boring configuration debt that compounds. Your booking engine runs on a separate domain, and nobody set up cross-domain tracking, so every reservation shows up as a referral from your booking provider and your marketing gets zero credit for revenue it created. PayPal and Stripe aren’t in your referral exclusions, so your "top converting source" is literally a payment processor. Facebook, facebook, and FACEBOOK are three different sources in your report, because nobody agreed on a naming convention. Your front desk, your marketing team, and your agency all count as guests, which on smaller properties inflates traffic by 5 to 15%. ChatGPT and Perplexity are already sending you visitors: 63% of sites now get at least one AI visitor, and those visitors read about twice as many pages as organic search visitors before they act. You’re not tracking any of it.
Fixing analytics is the least glamorous thing we do for a client. It’s also the first. Before a redesign, before a campaign, before a single dollar of new spend. Everything downstream inherits the error, so the error is where you start. We wrote down how to do it properly. Together with Cogwheel Marketing, we put the whole thing into a Best Practices Guide. Not theory. The exact settings, the exact mistakes, the exact fixes, written for hotels. You can’t fix what you can’t see. And right now, most hotels are flying blind.
Cogwheel Marketing and Zerf, GA4 Best Practices Guide for hotels. Search Engine Journal, common GA4 configuration mistakes. Gartner, the cost of poor data quality. The Net Revenue, GA4 and booking-engine tracking. OtelCiro GA4 guide.
