GA4 Mastery: The Metrics That Actually Predict Growth

Most businesses are drowning in GA4 data but starving for insight. They check sessions, bounce rates, and page views — and conclude their analytics are "fine". Meanwhile, the metrics that actually correlate with revenue growth sit buried in reports they've never opened. This guide cuts through the noise and focuses on what matters.

Stop Tracking These (They Don't Predict Revenue)

Before we talk about what to track, let's kill the metrics that consume reporting time without driving decisions:

Total sessions: Meaningless without segmentation by source, intent, and behaviour
Average session duration: Easily distorted by a small number of highly engaged outliers
Bounce rate (as traditionally defined): GA4 replaced this with Engagement Rate, which is more meaningful
Social media referral traffic: Usually tiny and hard to attribute correctly

The Metrics That Actually Matter

Rule of thumb: If a metric doesn't directly influence a decision you'd make, stop tracking it. Analytics should drive action — not fill dashboards.

Engaged Sessions per User: Sessions lasting 10+ seconds with 2+ page views or a conversion. This is GA4's replacement for bounce rate and far more meaningful.
Key Event Rate (formerly Conversion Rate): The percentage of sessions that result in a meaningful action. Set up micro-conversions (email signups, quote requests, video plays) in addition to purchases.
Revenue per User: Average revenue per unique visitor. This is the north star metric for e-commerce — more useful than average order value alone.
First Visit Conversion Rate: How often new visitors convert vs returning visitors. A huge gap signals onboarding or trust problems.

Setting Up GA4 for Revenue Intelligence

Out-of-the-box GA4 is a starting point, not a finished product. Configure these to unlock real intelligence:

Custom Events: Track micro-conversions — form starts, scroll depth milestones, video completion, file downloads
Enhanced Measurement: Enable all enhanced measurement options in GA4 settings — scroll tracking, outbound clicks, site search, form interactions
Audiences: Create audiences based on behaviour (visited pricing page but didn't convert, added to cart but abandoned) for remarketing and analysis
Channel Groups: Customise default channel groupings to match how your business actually acquires customers

Attribution in GA4: Data-Driven Is Best

Switch your attribution model to "data-driven" in GA4's Advertising section if you haven't already. This uses Google's machine learning to assign credit across the customer journey rather than defaulting to last-click. For most businesses, this reveals that top-of-funnel channels (SEO, social, YouTube) are significantly undervalued by last-click models.

GA4 Predictive Metrics: Hidden Gems

GA4 offers predictive metrics powered by Google's AI — but few businesses use them. These are genuinely useful:

Purchase Probability: The likelihood that a user who was active in the last 28 days will make a purchase in the next 7 days. Build remarketing audiences from high-probability users.
Churn Probability: The likelihood that an active user won't return. Use to trigger retention campaigns before you lose them.
Predicted Revenue: Revenue you can expect from users in the next 28 days. Useful for forecasting and budget allocation.

The insight principle: GA4 data has value only when it changes behaviour. Before opening any report, ask: "What decision am I trying to make?" If you don't have an answer, you're doing analytics theatre — not analytics.