Google Analytics 4 Setup Guide: Track What Actually Matters

We audited a client's GA4 setup last month and found 47 custom events being tracked. Forty-seven. When we asked which ones they actually used for decisions, the marketing manager paused, then said: "Honestly? I look at page views and bounce rate." That's the GA4 problem in a nutshell -- the tool can track everything, so people track everything, and then use none of it.

Over the past two years, we've set up or rebuilt GA4 for about 35 client websites. The setups that actually drive decisions have 8-12 carefully chosen events, 3-5 conversions, and two custom reports. That's it. Everything else is noise.

GA4 vs. Universal Analytics: What Actually Changed

If you're still thinking in Universal Analytics terms, the shift to GA4 isn't just a new interface -- it's a fundamentally different data model.

The Session-Based vs. Event-Based Difference

Universal Analytics was built around sessions and pageviews. Every interaction was tied to a session, and the session was the primary unit of analysis.

GA4 is built around events. Every interaction -- a pageview, a button click, a scroll, a purchase -- is an event. Sessions still exist, but they're derived from events rather than being the foundation.

What this means practically:

There is no "bounce rate" in the traditional sense. GA4 uses "engagement rate" (sessions lasting longer than 10 seconds, having 2+ pageviews, or triggering a conversion event). This is actually more useful -- a user who reads your entire blog post in 8 minutes and leaves isn't a "bounce" in any meaningful sense.
Pageviews are just another event (page_view), not a special category.
You can track user journeys across platforms (web + app) in a single property.
Attribution modelling is data-driven by default, not last-click.

What You Lose

Truthfully, some things were easier in Universal Analytics. Goals were simpler to set up than conversions. The interface was more intuitive. Custom reports were more straightforward. And if you relied on Content Grouping, that requires manual setup in GA4.

But UA is gone. Dwelling on it doesn't help. Let's set up GA4 properly.

Step 1: Property Configuration (15 Minutes)

Assuming you have a Google Analytics account, create a new GA4 property or configure your existing one.

Essential Settings

Admin > Data Streams > Your Web Stream:

Enable Enhanced Measurement. This auto-tracks: page views, scrolls (90% depth), outbound clicks, site search, video engagement (YouTube embeds), and file downloads. That's six useful events you don't need to code manually.
Set your default reporting identity to "Blended" (uses User-ID, Google signals, device ID, and modelling).

Admin > Data Settings > Data Retention:

Change from the default 2 months to 14 months. The default is absurd for any meaningful analysis. This only affects Explorations (custom reports), not standard reports.

Admin > Data Settings > Data Collection:

Enable Google Signals if you want cross-device tracking and demographic data. Be aware this has GDPR implications -- ensure your cookie consent covers it.

Admin > Property Settings > Reporting Identity:

"Blended" for most businesses. "Observed" if you don't want modelled data.

Internal Traffic Filter

Go to Admin > Data Streams > Configure Tag Settings > Define Internal Traffic. Add your office IP address(es). Then go to Admin > Data Settings > Data Filters and activate the internal traffic filter. We've seen clients pollute months of data because they forgot this step -- their own team's 200 daily visits were skewing every metric.

Step 2: Events That Actually Matter (30 Minutes)

GA4 has four categories of events: automatically collected, enhanced measurement, recommended, and custom. Here's what you need from each.

Keep From Enhanced Measurement (Already Tracking)

page_view -- obvious
scroll -- tells you if content is being consumed
click (outbound) -- where are users going when they leave?
file_download -- are people grabbing your PDFs, proposals, resources?
view_search_results -- what are people searching for on your site? This is free market research.

Recommended Events to Add

For lead generation sites (service businesses, agencies, consultants):

generate_lead -- fires when a contact form is submitted successfully
sign_up -- newsletter subscription or account creation
begin_checkout -- if you have any kind of quote or booking flow, track when it starts

For e-commerce sites, add the full e-commerce event set:

view_item, add_to_cart, begin_checkout, purchase
These feed into GA4's built-in e-commerce reports and are worth the setup time

Custom Events Worth Creating

Based on what we've found useful across client accounts:

cta_click -- track clicks on your primary calls to action. Use parameters to identify which CTA (cta_text, cta_location).
form_start -- fires when a user begins filling out a form. Compare with generate_lead to calculate form abandonment rate.
phone_click -- track tel: link clicks on mobile. For many service businesses, 40-60% of conversions come through phone calls.
scroll_25 / scroll_50 / scroll_75 -- enhanced measurement only tracks 90% scroll. Adding 25/50/75 thresholds tells you where people drop off. We use these on long-form landing pages.

Events You Don't Need

Stop tracking:

Every single button click. Track CTAs, not UI navigation.
Mouse hover events. They tell you nothing actionable.
Timer events ("user spent 30 seconds on page"). GA4 already measures engagement time.
Social share button clicks (unless social is a primary business channel -- for most, it's not).

Step 3: Mark Your Conversions (5 Minutes)

In GA4, any event can be marked as a conversion (called "key events" in newer interface versions). Go to Admin > Events, find your event, and toggle the "Mark as key event" switch.

The Conversions That Matter

Tier 1 -- Business outcomes:

purchase (e-commerce)
generate_lead (contact form submissions)
phone_click (mobile call taps)

Tier 2 -- High-intent actions:

begin_checkout or form_start (intent without completion -- useful for funnel analysis)
sign_up (newsletter/account)

Tier 3 -- Engagement signals:

file_download (especially for B2B where whitepapers/proposals are part of the funnel)

That's 4-6 conversions for most businesses. If you have more than 8, you're diluting the meaning of "conversion."

Step 4: Set Up Google Tag Manager (45 Minutes)

You can implement GA4 events through gtag.js directly or through Google Tag Manager (GTM). Use GTM. It's more work upfront but saves enormous time later. You won't need a developer every time you want to track something new.

The Essential GTM Setup

GA4 Configuration Tag -- fires on all pages, sends your Measurement ID
Contact Form Submission Trigger -- depends on your form setup. For most, this is a custom event triggered by the form's success callback, a thank-you page URL, or a dataLayer push.
CTA Click Triggers -- use Click Element matching against CSS classes or IDs on your primary CTAs.
Phone Click Trigger -- Click URL contains "tel:"

The DataLayer Approach

For reliable tracking, push events to the dataLayer from your site code:

dataLayer.push({
  'event': 'generate_lead',
  'form_name': 'contact_main',
  'form_location': 'homepage'
});

Then create a GTM trigger that listens for the custom event generate_lead. This is more reliable than DOM-based triggers because it fires exactly when your form logic confirms success, not when the submit button is clicked (which could fail validation).

We implement dataLayer pushes for every client site we build. It adds maybe 30 minutes of development time and saves hours of debugging flaky triggers later.

Step 5: Custom Reports That Drive Decisions

GA4's standard reports are a starting point, but the real value is in Explorations. Here are the three custom reports we build for every client:

1. Conversion Path Report

Type: Funnel Exploration

Steps: Landing Page View > CTA Click or Form Start > Form Submission or Purchase

This shows you where people drop off. If 1,000 people land on your services page, 200 click the CTA, and 15 submit the form, you know your landing page-to-CTA rate is 20% (decent) but your form completion rate is 7.5% (needs work -- the form is probably too long or asking for too much).

2. Content Performance Report

Type: Free Form Exploration

Dimensions: Page path, Session source/medium Metrics: Views, Engagement rate, Average engagement time, Conversions

This tells you which content is actually driving business results, not just traffic. We had a client whose most-visited blog post (15,000 monthly views) generated zero conversions. Their fifth-most-visited post (2,100 views) generated 40% of all form submissions. Without this report, they'd have kept pouring resources into the wrong content.

3. Traffic Quality Report

Type: Free Form Exploration

Dimensions: Session source/medium, Session campaign Metrics: Sessions, Engagement rate, Conversions, Conversion rate

Sort by conversion rate, not sessions. This exposes which channels send tyre-kickers versus actual prospects. One client discovered their LinkedIn organic traffic converted at 8.2% while their Google Ads traffic converted at 0.9%. They shifted $2,000/month from Google Ads to LinkedIn content creation and saw total conversions increase by 35%.

Common GA4 Mistakes We Fix Regularly

Duplicate page_view events. If you have GA4's Enhanced Measurement enabled AND a manual page_view tag in GTM, you're double-counting every pageview. Pick one.

Not filtering internal traffic. Your team visits your site daily. If you have 10 employees checking the site 5 times a day, that's 1,500 fake sessions per month.

Cross-domain tracking not configured. If your checkout or booking system is on a different domain, you need cross-domain tracking configured in your GA4 tag settings. Otherwise, every cross-domain visit looks like a new user from a referral.

Not connecting Google Search Console. Free integration, takes 2 minutes, gives you search query data inside GA4. There's no reason not to do this.

Using default channel groupings without review. GA4's automatic channel grouping puts traffic into categories like Organic Search, Direct, Social, etc. But it often miscategorises paid traffic if UTM parameters aren't consistent. Audit your channel groupings quarterly.

The 80/20 of GA4

After setting up 35+ GA4 properties, here's the honest truth: most businesses need to check three things weekly, and everything else is monthly or quarterly.

Weekly check (5 minutes):

Total conversions vs. previous week
Top traffic sources by conversion rate
Any anomalies in traffic volume

Monthly review (30 minutes):

Content performance report
Funnel drop-off analysis
Campaign performance comparison

Quarterly deep dive (2 hours):

Full channel quality assessment
Conversion path analysis
Data hygiene audit (duplicate events, filter accuracy, tag firing correctly)

GA4 is powerful. It's also easy to drown in. Set up the events that tie to business outcomes, build the three reports that drive decisions, and check them on a schedule. Ignore the rest until you have a specific question that requires deeper analysis. The clients who get the most value from analytics aren't the ones with the most complex setups -- they're the ones who actually look at their data and act on it.