Google Analytics 4 practical guide for marketers switching from Universal Analytics

Executive overview

Universal Analytics was shut down because it was built in a pre-mobile, pre-GDPR era — cookies, privacy violations, and no path to modernise. GA4 is a fundamentally different product, not an upgrade.

The core unlock is combining customised reports, the library, and funnel explorations. Each piece exists in isolation but together they give teams a shareable, actionable analytics workflow.

GA4's real power is unlocked only when custom reports, collections, and funnel explorations are combined — not used in isolation.

Why GA4 replaced Universal Analytics

• Universal Analytics stored excessive cookie data; GDPR made this untenable
• GA4 is privacy-first by design, built for a fragmented, multi-device world
• Adoption: ~50% of top 10K/100K sites use GA4; many stragglers remain on old code
• Constant product updates — Google ships changes daily
• Free BigQuery export, native Search Console and Google Ads integration

Key advantages of GA4

• Predictive analytics and data-gap modelling fill in blind spots from blocked cookies
• Attribution modelling across channels, built into the product
• Monetisation reports that go beyond e-commerce — tracks any revenue pathway
• Fresh start: fix historical data mistakes that were permanent in Universal Analytics
• Independent scoring across 40+ analytics tools ranked GA4 highest overall

Customising and saving reports

• Default acquisition report uses default channel grouping — not source/medium
• Use "Customise report" to set source/medium as the default dimension permanently
• "Save to current report" changes the view for all users on the property — use with caution in agency accounts
• "Save as new report" creates a personal copy; add it to a collection to make it navigable
• The library is where all saved reports live; reports must be added to a published collection to appear in the nav

Collections and the library

• Reports are organised into collections (e.g. Lifecycle, Users, Business Objectives)
• Business Objectives collection uses plain-language names; identical reports to Lifecycle but easier to navigate
• Agencies: create separate collections per client or product line — keeps custom reports contained
• Publish a collection to make it visible to all property users; unpublished collections are only visible to you
• Saved reports in a collection are visible and interactive for everyone with property access

Funnel explorations

• Explorations are private by default — only the creator can view or edit
• Use "Funnel exploration" type to map user journeys using page location as the step condition
• Works with basic G tag implementation — no Tag Manager or e-commerce setup required
• Name each step explicitly (e.g. "Offer page", "Cart page", "TY/Upsell page") for readability
• Completion rates show step-to-step conversion; overall rates show conversion from the total starting cohort

Adding forecast benchmarks to funnels

• Add a "forecast" column to each funnel step with expected conversion range (e.g. 30–40% for opt-in, 8–12% for offer-to-cart)
• Benchmarks make underperforming steps immediately visible — no mental arithmetic required
• Typical benchmarks: opt-in 30–40%, offer-to-cart 8–12%, cart conversion 40–50%
• Break down by session source/medium/campaign to isolate which traffic source is underperforming
• Expand breakdown from top 5 to up to 15 sources for deeper analysis

Sharing explorations as reports

• Explorations can be shared read-only — but recipients cannot change dates, breakdowns, or filters
• Use "Save as report in library" to convert an exploration into a fully interactive shared report
• Once saved to the library, add it to a collection — it immediately becomes available to all property users
• New (unreleased at time of recording): edit dimensions directly on a library report without going back to the exploration

Implementation and migration notes

• Leaving old Universal Analytics code on the site is not penalised — Google ignores UA hits
• Duplicate GA4 code (native + converted UA) creates two properties; one is inferior — clean it up
• Tag Manager is strongly recommended over direct G tag if you need custom events, scroll tracking, or custom dimensions
• Use the G tag directly only if GA4 is the sole tool and no customisation is needed
• Bounce rate is back in GA4 (added post-launch) but engagement rate is a more actionable replacement metric

Transitioning stakeholders from Universal Analytics

• Bounce rate has been absent for 11+ months — stakeholder conversations should already have happened
• Frame new metrics as better indicators, not replacements — show what action they enable
• Focus on metrics tied to decisions: conversion rates, traffic source performance, funnel drop-off
• HubSpot Analytics and GA4 are complementary — GA4 fills gaps HubSpot cannot cover