Measuring Direct Bookings: A GA4 Starting Point for Operators

The short answer: Most serviced accommodation operators have GA4 installed and have never opened it, because the default reports don’t answer the one question that matters: is my own website driving bookings, and is that improving. Fix that by marking your enquiry or booking event as a key event, checking your channel grouping is accurate, and building one saved exploration that compares your own channels against OTA-referred traffic. Fifteen minutes a month is enough once it’s set up properly.

If you’ve never opened Google Analytics 4 beyond installing it, you’re not unusual. GA4 replaced the older, more intuitive Universal Analytics in 2023, and the new interface buries the one report most operators actually want behind menus built for e-commerce teams, not accommodation businesses.

This is not a complete analytics course. It’s the minimum working setup to answer a single question: is your own website generating direct bookings, and is that number moving in the right direction. Everything below assumes GA4 is already installed on your site. If you’re not sure, the first step below tells you how to check.

Why this matters more than it used to

Every operator watching OTA commission eat into margin has heard some version of “build direct bookings.” Fewer have a way to prove whether it’s working. Without measurement, “direct bookings are growing” is a feeling, not a fact, and marketing budget gets allocated on feelings.

Relocationapartments.com, a retired W4B-built site, is a useful example of what consistent measurement over time actually reveals. Across a twelve-month reporting period the site logged 236,471 visits and 856,856 pageviews, averaging around 19,700 visits a month, with 9,104 of those visits reaching the enquiry form page specifically. None of that is impressive as a single number. What makes it useful is that it’s a twelve-month trend, pulled from the site’s own server logs, that let the operator see which months and which pages were actually producing enquiry traffic rather than just visits. That’s the habit this post is trying to build, not the specific tool.

The reason GA4 specifically causes trouble is that it replaced the older, more familiar Universal Analytics in 2023 with a fundamentally different data model built around events rather than pageviews and sessions in the old sense. Operators who learned Universal Analytics years ago open GA4, don’t recognise the interface, and quietly stop looking. The setup below is deliberately narrow. It is not a tour of GA4’s full feature set, most of which an SA operator will never need. It is the smallest set of steps that gets you from an installed-but-ignored account to one working report you’ll actually open every month.

Step 1: Confirm GA4 is actually collecting data

Before trusting anything else, check the basics. Open GA4’s Realtime report, then open your own website in a second tab or your phone and browse a few pages. If your own visit shows up in Realtime within a minute or two, tracking is working. If it doesn’t, stop here and fix the tracking code before reading another report, because every number below is meaningless without it.

Step 2: Mark your enquiry or booking event as a key event

This is the single most important setup step, and the one most operators skip. By default, GA4 tracks page views and generic engagement, but it has no idea which of those page views actually mattered to your business.

Find the event that fires when a guest submits an enquiry form or completes a booking, and mark it as a key event (GA4’s current term for what used to be called a conversion) in Admin settings. Once this is done, every acquisition report in GA4 can show you not just traffic by channel, but which channels actually produced enquiries, which is the report that turns analytics from a vanity number into a decision-making tool.

Step 3: Check your channel grouping is telling the truth

GA4 automatically groups traffic into channels: Organic Search, Direct, Paid Search, Email, Social, Referral, and so on. The defaults are usually reasonable, but two common failures are worth checking specifically:

• Branded paid search showing as Direct. If a guest searches your property name and clicks a paid ad, poorly tagged campaigns can sometimes get miscategorised as Direct traffic, which inflates your direct-booking number without anything having actually changed.
• Email links without tracking parameters showing as Direct. Links in newsletters or post-stay email sequences need UTM parameters to be correctly attributed to Email rather than falling into Direct or Referral.

Open the Traffic Acquisition report, sorted by sessions, and sanity-check the top few channels against what you know is actually happening. If something looks implausibly high or low, the tagging is usually the cause.

Step 4: Build one exploration worth checking

GA4’s standard reports are built for general e-commerce, not for the specific question an SA operator has: how is my own site performing against the OTA-referred and search traffic I don’t control.

In the Explore section, build a free-form exploration with channel as a dimension, and sessions plus your key event as metrics, over a rolling 90-day window. Save it. This single view answers the practical question every month: is direct-channel traffic (Direct, Organic Search, Email) holding or growing its share of enquiries relative to everything else.

If you want the wider context for what “everything else” should include, our guide to reducing OTA dependency covers the levers that actually move this number, and why your serviced apartment website isn’t getting bookings works through the usual reasons the traffic you do measure isn’t converting.

Step 5: Review monthly, not daily

SA booking volumes are naturally lumpy. A single busy weekend or a single group enquiry can swing a week’s numbers in a way that means nothing about the underlying trend. Checking GA4 daily mostly trains you to react to noise.

Put a fifteen-minute slot in the calendar once a month, open the saved exploration from step 4, and compare it against the previous month and the same month a year earlier if you have the history. That’s the full review. Anything more frequent is usually procrastination dressed up as diligence.

Common mistakes that make the numbers lie

Even with the setup above, a handful of habits quietly corrupt GA4 data for SA operators specifically.

Not excluding your own team’s traffic. If your booking or property management staff browse the live site regularly from the office, without a filter excluding your office IP or a browser extension, their visits sit inside your real guest numbers. On a smaller site this can meaningfully inflate session counts and distort which pages look popular.

Treating cross-domain journeys as two separate visits. If your booking engine lives on a different domain to your main site, and cross-domain tracking isn’t configured, GA4 sees a guest who browses your site and then books on the engine’s domain as two unrelated sessions rather than one journey. This is one of the most common reasons the enquiry-to-booking funnel looks worse in GA4 than it actually is.

Reading small-sample explorations as trends. Most SA operators run one to a handful of properties, not the volume a large hotel chain sees. A 90-day exploration with 40 total conversions can swing 20% between two consecutive months from ordinary variation. Look for a consistent direction across two or three consecutive months before acting on a change, not a single month’s number.

Comparing the wrong periods. SA demand is seasonal. Comparing October to September tells you about seasonality, not about whether your marketing is working. Compare each month against the same month a year earlier wherever you have the history, and use the rolling 90-day view for the underlying trend in between.

If you manage more than one property

Multi-property operators can add a custom dimension or use separate landing pages per property so GA4 can break results down by property rather than only by channel. Without this, a strong month for one property and a weak month for another simply average out into a flat-looking total, hiding exactly the information a multi-property operator most needs: which property’s marketing is working and which one needs attention.

The simplest version of this, before touching custom dimensions, is making sure each property has its own dedicated landing page with its own URL, which the free-form exploration in step 4 can then filter or break down by page path.

What GA4 alone won’t tell you

GA4 tells you what happened on the website. It won’t tell you why an enquiry didn’t convert to a booking, whether a guest who left a review came from a direct booking or an OTA one, or whether a slow page cost you the visit in the first place. For that context, pair GA4 with your booking system’s own enquiry-to-booking conversion rate, and if speed is a live suspicion, see why your booking site is slow for how to check that specifically.

If you’ve never built a proper measurement habit and aren’t sure where GA4 fits against everything else worth doing, the cheat sheet sets out the order operators get the most value from, measurement included. It’s a faster starting point than trying to learn GA4 in isolation, and downloading it is free.

Frequently Asked Questions

Do I need GA4 if I already track bookings in my PMS or channel manager?

Your PMS tells you what was booked. GA4 tells you what happened before the booking, which pages a guest viewed, which channel brought them, and whether they dropped off at the enquiry form. The two together answer more than either alone, and neither replaces the other.

How long does GA4 take to become reliable after installation?

Basic event tracking works within minutes, visible in Realtime reports. But trend reports need a running history to be useful, and most operators find a rolling 90-day view is the shortest window that smooths out normal week-to-week noise.

What’s the single most useful GA4 report for a serviced accommodation operator?

A saved exploration comparing sessions and your enquiry key event by channel, over a 90-day rolling window. It’s the one view that directly answers whether direct-channel effort is working, and it doesn’t exist by default, which is exactly why most operators who’ve opened GA4 once gave up on it.

Is GA4 enough on its own, or do I need Google Search Console too?

Both, for different questions. GA4 tells you what visitors did once they arrived. Search Console tells you what they searched to find you and where you rank for it. If your enquiries are thin despite decent traffic, GA4 will show that; if traffic itself is thin, Search Console is where you’ll see why.