Google Analytics 4

Google Analytics 4 is the most significant evolution in web analytics in over a decade — a complete rethinking of how businesses measure user behaviour, track conversions, and attribute marketing performance. For any business serious about making data-driven decisions, GA4 is no longer optional. It is the foundation on which every other marketing measurement effort is built.

What Is Google Analytics 4?

Google Analytics 4 (GA4) is Google's current-generation analytics platform, replacing Universal Analytics (UA), which was officially sunset in July 2023. While Universal Analytics was built around sessions and pageviews — a model designed for a simpler, desktop-first web — GA4 is built on an event-based data model that tracks every user interaction as a discrete event. This shift makes GA4 fundamentally more flexible, more powerful, and better suited to the modern, cross-device customer journey.

Unlike Universal Analytics, which treated a "session" as the primary unit of measurement, GA4 treats every action — a page view, a button click, a scroll, a video play, a form submission — as an event with its own parameters. This means you can capture far richer behavioural data without being constrained by the session-based logic of older analytics platforms.

Key Differences from Universal Analytics

Event-based vs. session-based — GA4 records individual user interactions as events; UA grouped them into sessions
Cross-device tracking — GA4 uses a unified identity framework combining Google signals, User IDs, and device IDs to stitch together journeys across phones, tablets, and desktops
Privacy-first architecture — GA4 is designed for a cookieless future, with modelled data filling gaps where consent has not been granted
Built-in BigQuery export — All GA4 properties (including free accounts) can export raw event data to BigQuery for advanced analysis
Data-driven attribution as default — GA4 uses machine learning-based attribution rather than last-click by default
Predictive audiences — GA4 can predict which users are likely to purchase or churn, enabling proactive remarketing

The Event-Based Data Model

Understanding GA4's event model is essential to using the platform effectively. Every interaction tracked in GA4 is an event — a named action with optional parameters that add context.

GA4 automatically collects a core set of events out of the box, including:

page_view — triggered on every page load
session_start — triggered when a new session begins
first_visit — triggered the first time a user visits your site
scroll — triggered when a user reaches 90% of a page
click — triggered on outbound link clicks
file_download — triggered when a user downloads a file
video_start, video_progress, video_complete — triggered for embedded YouTube videos
form_start, form_submit — available via enhanced measurement

Beyond these automatic events, businesses can define custom events to track any interaction specific to their site — a specific button click, a product filter use, a chat widget open, or a loyalty programme sign-up.

Key Events (Conversions)

In GA4, any event can be designated as a Key Event (previously called a Conversion). This replaces UA's goal system with a more flexible framework — rather than configuring a fixed number of goal slots, you can mark as many events as needed as Key Events and GA4 will report on them prominently throughout the interface.

Common Key Events for most businesses include:

Form submissions and lead enquiries
Phone call initiations (via call tracking integration)
Purchase completions
Newsletter sign-ups
Account registrations
Quote requests

Setting Up GA4: A Practical Framework

A well-configured GA4 property is the result of deliberate planning, not just installing a tracking snippet. The following framework reflects best practices for businesses setting up or auditing their GA4 implementation.

1. Define Your Measurement Plan

Before touching a single tag, document what you actually need to measure. Start with your business objectives — what does success look like? — and work backwards to identify the events and Key Events that map to those outcomes. A measurement plan prevents bloated, inconsistent tracking and ensures your data answers real business questions.

2. Install via Google Tag Manager

While GA4 can be installed by pasting the gtag.js snippet directly into your site's code, the best practice is to deploy it through Google Tag Manager (GTM). GTM acts as a container for all your tracking tags — GA4, Google Ads conversion tags, Meta Pixel, LinkedIn Insight Tag, and more — managed from a single interface without requiring developer involvement for every change.

Key advantages of using GTM:

Deploy and update tags without touching site code
Test tags in a sandboxed preview environment before publishing
Manage triggers and variables centrally across all tags
Implement server-side tagging for improved data quality and privacy compliance
Maintain a version history of all tag changes

3. Configure Enhanced Measurement

GA4's Enhanced Measurement feature automatically tracks a suite of common interactions — scrolls, outbound clicks, site search, video engagement, and file downloads — without any additional code. Enable this in your GA4 property settings as a baseline, then layer custom events on top for interactions specific to your business.

4. Set Up Key Events

Navigate to Admin → Events in your GA4 property and toggle the "Mark as key event" switch for any event that represents a meaningful conversion action. For e-commerce, this includes purchase. For lead generation businesses, this typically includes form_submit, generate_lead, or a custom event like quote_request.

5. Link to Google Ads and Search Console

Connecting GA4 to your Google Ads account enables:

Importing GA4 Key Events directly into Google Ads as conversion actions
Activating GA4 audience segments for remarketing in Google Ads
Viewing Google Ads campaign performance within GA4 reports

Linking GA4 to Google Search Console surfaces organic search performance data — queries, impressions, clicks, and average position — directly inside GA4's reporting interface.

6. Enable BigQuery Export

For businesses that need granular, unsampled data or want to build custom dashboards and attribution models, enabling the free BigQuery export is a critical step. Raw event-level data exported to BigQuery can be queried with SQL, visualised in Looker Studio, or fed into machine learning models.

Attribution Modelling in GA4

Attribution is the practice of assigning credit to the marketing touchpoints that contributed to a conversion. It answers the fundamental question: which channels and campaigns are actually driving results?

GA4 has made data-driven attribution (DDA) the default model for all properties with sufficient data. This represents a significant departure from the last-click attribution that dominated analytics for years.

Understanding the Main Attribution Models

Last Click — 100% of conversion credit goes to the final touchpoint before conversion. Simple, but systematically undervalues upper-funnel channels like display and social that introduce customers to a brand.
First Click — 100% of credit goes to the first touchpoint. Useful for understanding what initiates the customer journey, but ignores everything that happens after.
Linear — Credit is distributed equally across all touchpoints in the conversion path. More balanced, but treats a brand awareness display impression the same as a high-intent search click.
Time Decay — More credit is given to touchpoints closer in time to the conversion. Useful for short sales cycles.
Position-Based — 40% of credit each to the first and last touchpoints, with the remaining 20% distributed across middle interactions.
Data-Driven Attribution — Uses machine learning to analyse your actual conversion data and assign credit based on the real contribution of each touchpoint. Adapts dynamically as user behaviour changes and requires a minimum data threshold to activate.

According to Fluxion Digital's 2025 attribution research, organisations that have invested in data-driven attribution infrastructure are seeing 30–40% more accurate attribution compared to those relying on fragmented tracking and last-click models.

The Limits of GA4 Attribution

No attribution model is perfect. GA4 attribution operates within the boundaries of what it can observe — it cannot account for offline interactions, word-of-mouth referrals, or touchpoints that occurred before a user consented to tracking. For businesses with complex, multi-channel marketing mixes, GA4 attribution should be treated as one input alongside incrementality testing and marketing mix modelling, rather than the definitive source of truth.

Google Tag Manager: The Engine Behind Accurate Tracking

While GA4 is the analytics platform, Google Tag Manager is the infrastructure layer that makes accurate, scalable tracking possible. Understanding how to use GTM effectively is as important as understanding GA4 itself.

GTM operates on three core concepts:

Tags — The tracking code snippets that fire when conditions are met (e.g., a GA4 event tag, a Google Ads conversion tag, a Meta Pixel)
Triggers — The conditions that cause a tag to fire (e.g., a form submission, a button click, a page view on a specific URL)
Variables — Dynamic values used by tags and triggers (e.g., the page URL, a product name from the data layer, a click element's CSS class)

The Data Layer

The data layer is a JavaScript object that acts as a structured communication channel between your website and GTM. When a user adds a product to their cart, your website's code pushes structured data — product name, price, category, quantity — into the data layer. GTM reads this data and passes it to GA4, Google Ads, and any other tags that need it.

Implementing a well-structured data layer is the single most important technical investment in any analytics setup. It decouples tracking logic from site code, making future changes far easier and ensuring data consistency across all platforms.

Server-Side Tagging

Traditional client-side tagging — where tracking code runs in the user's browser — is increasingly limited by ad blockers, browser privacy restrictions, and iOS tracking changes. Server-side tagging moves tag execution from the browser to a server environment you control, improving data quality, reducing page load impact, and giving you greater control over what data is shared with third-party platforms.

For businesses running significant paid media budgets, server-side tagging is becoming a best-practice standard rather than an advanced option.

Real-World Examples: Analytics Driving Business Results

Medical Clinic: From Vanity Metrics to 275% Conversion Growth

A medical clinic had been running Google Ads for months but was only tracking clicks on website elements — not the phone calls that represented their actual patient enquiries. Once proper call tracking was configured in GA4 and integrated with Google Ads, the true conversion picture emerged. By optimising campaigns around real conversion data rather than proxy metrics, the clinic achieved 275% conversion growth while reducing cost-per-conversion by 46%. The lesson: accurate conversion tracking does not just improve reporting — it fundamentally changes how campaigns are optimised.

E-Commerce Brand: Recovering Lost Revenue Through Funnel Analysis

GA4's funnel exploration reports allow businesses to visualise exactly where users drop off during a multi-step process — checkout, sign-up, onboarding. A mid-sized e-commerce brand used GA4's funnel analysis to identify that 68% of users were abandoning at the payment details step on mobile. Investigation revealed a form field validation issue that only appeared on certain Android devices. Fixing the bug — identified entirely through GA4 data — recovered an estimated 15% of previously lost revenue within the first month.

B2B SaaS: Attribution Rethink Unlocks Paid Social Investment

A B2B software company had historically underinvested in LinkedIn Ads because last-click attribution showed minimal direct conversions from the channel. After switching to data-driven attribution in GA4 and layering in incrementality testing, the team discovered that LinkedIn was consistently appearing as a critical early touchpoint for their highest-value enterprise deals — customers who went on to convert via branded search weeks later. Reallocating budget based on the fuller attribution picture increased pipeline by 22% within a quarter.

GA4 Reports: What to Actually Look At

GA4's reporting interface can feel overwhelming at first. Here are the most practically useful reports for most businesses:

Acquisition Reports

Traffic Acquisition — Shows where sessions are coming from by channel (Organic Search, Paid Search, Direct, Referral, Email, etc.)
User Acquisition — Shows what channel first brought each user to your site — useful for understanding which channels are best at introducing new audiences

Engagement Reports

Pages and Screens — Which pages are getting the most views and engagement
Events — A full list of all events being tracked, with counts
Key Events — Conversion performance by event type

Monetisation Reports

E-commerce Purchases — Revenue, transactions, and product performance for e-commerce sites
Purchase Journey — A funnel view from product view through to purchase completion

Exploration Reports

Funnel Exploration — Build custom conversion funnels to identify drop-off points
Path Exploration — Visualise the most common sequences of pages or events users move through
Segment Overlap — Compare how different audience segments overlap and convert

Best Practices for GA4 in 2025

Start with a measurement plan. Document your business objectives, the events that map to them, and the Key Events you'll track before writing a single line of code or configuring a single tag.
Deploy through Google Tag Manager. Avoid hardcoding GA4 tags into your site's source code. GTM gives you the flexibility to iterate quickly, test safely, and maintain clean version control.
Use the data layer for structured data. For e-commerce and any site with dynamic content, implement a data layer to pass structured, reliable data to GA4 and other platforms.
Mark the right things as Key Events. Only designate events as Key Events if they represent genuine business value. Over-marking inflates your conversion numbers and makes Smart Bidding in Google Ads less effective.
Link GA4 to Google Ads immediately. Importing GA4 Key Events into Google Ads is one of the highest-impact steps you can take. It gives Smart Bidding accurate signals to optimise against.
Switch to data-driven attribution. If your property has sufficient conversion volume, use data-driven attribution rather than last-click. It provides a materially more accurate picture of channel contribution.
Enable BigQuery export for scale. For businesses generating significant data volumes or needing custom reporting, the free BigQuery export is a game-changer. Build your dashboards in Looker Studio on top of unsampled, raw event data.
Audit your implementation regularly. GA4 setups drift over time as sites change. Schedule quarterly audits using GA4's DebugView and GTM's preview mode to verify that all Key Events are firing correctly and no duplicate tracking has crept in.
Respect consent requirements. GA4's consent mode allows you to adjust tracking behaviour based on user consent status, with modelled data filling gaps where cookies have been declined. Implementing consent mode correctly is both a legal requirement in most markets and a data quality measure.
Don't rely on GA4 alone. For businesses with significant marketing budgets, complement GA4 with platform-native reporting, incrementality testing, and — at enterprise scale — marketing mix modelling. No single tool tells the complete story.

Key Takeaways

Google Analytics 4 is a fundamentally more powerful analytics platform than its predecessor — but that power comes with complexity. Businesses that invest in a proper implementation, understand the event-based data model, and use GA4's attribution capabilities correctly will make materially better marketing decisions than those relying on incomplete data or outdated tracking setups.

GA4's event-based model is more flexible and more powerful than Universal Analytics' session-based approach
Accurate Key Event tracking is the foundation of every other measurement and optimisation effort — without it, all decisions are made in the dark
Google Tag Manager is the right way to deploy and manage GA4, and the data layer is the backbone of reliable event tracking
Data-driven attribution provides a more accurate view of channel contribution than last-click — and GA4 makes it the default
Real-world examples consistently show that improving analytics accuracy — not just increasing ad spend — is one of the highest-ROI actions a business can take
Regular audits, consent mode compliance, and BigQuery integration separate mature analytics setups from basic ones

For businesses serious about growth, GA4 is not just a reporting tool — it is the operating system for data-driven marketing.