GA4 Setup Checklist: Configure What You Can Verify

A reliable GA4 setup connects business questions to a tested collection path. Creating a property and seeing users in Realtime is only the beginning; ownership, consent, event definitions, referrals, integrations and validation determine whether the reports can support decisions.

Use this GA4 setup checklist as an implementation record. Every configured item should have an owner, reason and test result.

Quick answer: Create the account, property and web data stream under company ownership. Install one controlled Google tag route, configure consent and stream settings, define events and key events, link required products, filter internal traffic carefully, test with DebugView, and document the final measurement map.

1. Establish account ownership and property settings

The business—not an individual contractor’s private login—should retain appropriate administrative access. Use named users and the least privilege required for each role.

Check:

• account and property names;
• company-controlled administrator access;
• backup administrators;
• property time zone;
• reporting currency;
• industry and business settings where relevant;
• data-sharing choices;
• retention and privacy settings;
• change-history access;
• removal of obsolete users.

Time zone and currency affect interpretation, so confirm them before collecting production data. Changing settings later can complicate comparisons.

Document the property ID, data stream, Google tag ID and connected products. Never paste sensitive credentials into a general implementation document.

2. Create and configure the web data stream

A data stream sends website activity to the property. Confirm the production URL and a descriptive stream name.

Review enhanced measurement options individually. Page views, scrolls, outbound clicks, site search, video engagement, file downloads and form interactions may be available depending on the environment and current product behaviour. Enable only what you understand, and test whether the site’s architecture supports accurate collection.

For example, single-page applications need route-change validation. Site-search parameters must match the site. Automated form interactions may not represent successful submissions. Enhanced measurement can reduce implementation work, but it does not remove the need for definitions.

Also review:

• cross-domain measurement for genuine user journeys across owned domains;
• unwanted referrals from payment, booking or authentication providers;
• internal traffic definitions and filters;
• session settings where there is a documented reason;
• connected tags and legacy implementations.

3. Install one controlled tag path

GA4 can be installed through the Google tag, Google Tag Manager, a CMS integration or another supported platform. Choose a route that the team can own and test.

Avoid loading the same GA4 property independently through both hard-coded gtag.js and GTM unless the design explicitly prevents duplicates. Duplicate configurations can create repeated page views, events or inconsistent settings.

Verify the tag:

• appears on every intended template;
• loads before dependent events;
• uses the correct ID;
• does not load on staging into the production property unless planned;
• handles single-page navigation;
• preserves campaign parameters and cross-domain identity where required;
• respects consent choices;
• does not send personally identifiable information.

Use a test or development property or clear environment controls for complex implementations. Publishing unverified tags directly into production makes debugging and historical interpretation harder.

4. Configure consent and privacy behaviour

Analytics implementation is not a substitute for legal advice. Determine the consent and privacy requirements that apply to the organisation, then configure the consent management platform and Google tags accordingly.

For Google Consent Mode, the current implementation includes consent states such as analytics_storage, ad_storage, ad_user_data and ad_personalization. Default states must be set before measurement commands that depend on them, then updated when the user changes a choice. The exact defaults are a policy decision.

Test:

• first visit before interaction;
• accept all;
• reject all;
• granular choices;
• revocation;
• navigation after choice;
• return visit;
• regional behaviour if used;
• whether tags and requests match the chosen mode.

Do not assume a banner proves consent mode works. Verify the actual consent state and network behaviour with Tag Assistant and browser tools.

5. Design events, key events and ecommerce

Build a measurement plan before creating tags. Use recommended events where they fit and custom events only when necessary.

For each event, specify trigger, parameters, consent requirements, deduplication and test. GA4 ecommerce tracking setup must follow the recommended event and items structure, including required values such as transaction identifiers where applicable.

Mark only validated outcomes as key events. Review any Google Ads conversion created from them separately.

6. Link products and configure reporting

Connect products only when the business has an intended use and appropriate permissions. Common integrations can include Google Ads, Search Console and BigQuery export, depending on needs and eligibility.

For Google Ads:

• link the correct accounts;
• review personalisation and data-sharing choices;
• confirm auto-tagging and campaign parameters;
• decide which events become conversions;
• reconcile attribution and time-zone differences.

For reporting, verify acquisition, engagement, landing page, events, key events and ecommerce reports relevant to the measurement plan. Create custom dimensions only for parameters that stakeholders will use. High-cardinality or unstable values can make reports harder to interpret.

Avoid building a dashboard before definitions are agreed. A polished report can make incorrect data look more authoritative.

7. Validate, document and monitor

Complete the GA4 configuration checklist with controlled tests:

• Realtime and DebugView receive expected events;
• event parameters contain expected values;
• events do not duplicate on refresh or route change;
• internal traffic handling works as intended;
• cross-domain journeys preserve source and session behaviour;
• referrals from payment or booking systems are handled correctly;
• ecommerce totals reconcile with the commerce platform within known differences;
• CRM submissions match tracked leads;
• consent states produce expected behaviour;
• standard reports populate after processing.

Create a verification record and a data dictionary. Monitor sudden drops, spikes, (not set), unassigned traffic, new event names and changes to key-event counts. Review the change history when something shifts unexpectedly.

If the GA4 account setup needs to connect tags, consent, advertising and sales outcomes, Growthjunction’s analytics and tracking service focuses on a verified signal path rather than a larger dashboard.