Analytics and tracking
GA4 booking engine: what a hotel should track
Four events on the booking engine and the guest's source. From them a manager sees where guests drop off: at the rate, at the details or at payment. 64% of booking pages in our check have no tag at all.
Short answer
On the booking engine GA4 needs four events: view_item_list, add_to_cart, begin_checkout and purchase with value, plus the guest's source. The gap between steps shows where guests drop off. In our check of 84 Croatian hotels, 23 of 36 booking pages had no analytics tag at all.
- Four events on the engine and the guest's source. Everything else is extra.
- The drop between steps shows where guests leave: rate, details or payment.
- 14 of 22 hotels that measure the website do not measure the booking engine.
AI illustration
On the booking engine GA4 needs four events and the guest's source. Nothing more to start with. From them a manager sees where guests drop off and which channel brings bookings, not just visits.
Four events on the engine
- view_item_list - available rooms for the chosen dates
- add_to_cart - chosen room and rate
- begin_checkout - guest details
- purchase - confirmed booking, with transaction_id, value and currency
On the hotel website, also book_now_click, the click on Book now. I track phone and email clicks separately, as micro-conversions.
Names differ from engine to engine. What matters is one event per step and a purchase that arrives once.
How many hotels have it
In October 2026 we checked 84 Croatian 4 and 5 star hotels. We checked the booking page for 36 of them. 23 of 36 had no analytics tag at all. Of 22 hotels that measure the website, 14 do not measure the booking engine. 20 websites still carry old Universal Analytics code, which stopped collecting data in July 2023.
We looked before cookie consent, so these figures are a lower bound. The method is in the direct booking study.
How to read the steps
An example month:
| Step | Count | From previous |
|---|---|---|
| Sessions on the engine | 1,000 | |
| begin_checkout | 80 | 8% |
| purchase | 20 | 25% |
| Revenue at 400 EUR per booking | 8,000 EUR |
Do not compare these numbers with another hotel. Compare them with the same hotel, in the same part of the season.
What a drop means:
- many room views, few selections: rate, availability or minimum stay
- many selections, few checkouts: the step is complicated or slow on mobile
- many checkouts, few confirmations: payment, deposit or trust
Look at mobile separately from desktop first.
The guest's source
Every booking needs a source: Google Ads, Meta, organic, newsletter, direct. Without cross-domain measurement between the website and the engine the source is lost and everything looks like a referral from the hotel's own website. How to set this up on Phobs is in the article Phobs and GA4.
Sources
- Nikton Hospitality: direct bookings in Croatian hotels 2026.
- Google Analytics: recommended events for online sales.
Next step
Continue with the booking engine checklist. If you want someone to look at your engine and GA4, ask for a hotel audit.
Related: case study Meneghetti Wine Hotel & Winery. Campaigns, the booking engine and direct booking results, June 2026 vs June 2025, source GA4.
Frequently asked questions
Which events does a hotel need in GA4?
Four on the engine: view_item_list for available rooms, add_to_cart for the chosen room and rate, begin_checkout for guest details and purchase for the confirmed booking, with transaction_id, value and currency. On the website, also book_now_click.
Where do guests usually drop off?
It depends on the hotel, which is why you measure. Many room views and few selections usually means rate or availability. Many selections and few checkouts means a complicated step. Many checkouts and few confirmations means payment or trust.
Is 8% reaching begin_checkout a good result?
That is only the example in the article. The comparison only holds for the same hotel over time, in the same part of the season. Another hotel with another market and rate has another ratio.
Why does GA4 show fewer bookings than the engine?
Some guests refuse cookies, so GA4 sees fewer than the engine. That is normal. More purchases in GA4 than bookings in the engine means the purchase is counted twice.
Nikton Hospitality. A digital marketing agency for hotels.
About Nikton Hospitality and how we workRelated articles
Analytics and tracking
Phobs and GA4: one booking, one purchase
Phobs runs on its own domain. Without cross-domain measurement GA4 loses the guest's source, and the purchase is easily counted twice. On 15 of 21 Phobs booking pages we found neither GA4 nor GTM.
Analytics and tracking
GA4 events for the hotel booking funnel
A measurement plan template for a hotel booking engine: events, triggers, parameters and validation. Includes an abandonment calculation example and the limits of what it shows.
Booking engine
Booking engine, channel manager and Phobs: who sends the rate
The booking engine is where the guest completes a reservation for the hotel. The channel manager sends rates and availability to OTAs. Phobs still needs its role confirmed.