Content Moved? Use Search to Locate
Close-up of a social media marketing document on a desk with a pen and notebook.

GA4 Source Group Is Retroactive: Audit Traffic Before Changing Spend

BySplinternet Marketing

A higher or lower share for a social platform or AI referral in GA4 is not, by itself, proof that campaign efficiency changed. Google’s June 11, 2026 update added the Source Group dimension, revised some Source Platform classifications, and populated Source Group retroactively. Historical tables can therefore look different even when visits, ad spend, leads, orders, or revenue remain unchanged.

That creates a practical reporting risk. A business owner may reduce a channel that appears weaker, an agency may defend a different ROAS calculation, or a marketing lead may treat newly grouped AI referrals as a sudden acquisition trend. Before making a major budget decision, compare the same date range, dimensions, metrics, attribution settings, and reporting scope.

Need help checking this on your WordPress, Google Ads, Analytics, local SEO, or website setup? Splinternet Marketing can review the issue and help you prioritize the next fix.

Do not treat GA4 source dimensions as interchangeable

Source identifies the specific referring platform or location, such as Google, Facebook, or a newsletter name. Medium describes the traffic type, such as organic, cpc, referral, social, or email. Campaign identifies the marketing initiative.

Source Platform provides a broader platform classification. Default Channel Group organizes traffic into broad channel categories using source, medium, and channel-grouping rules. Source Group is intended to consolidate inconsistent source values into a cleaner platform-level reporting value.

For example, historical data may contain separate source values such as “facebook,” “fb,” or other variations. After retroactive grouping, those rows may appear under a more consistent Source Group value. Google’s release documentation also describes grouping for emerging AI sources, including ChatGPT and Perplexity.

The reported share assigned to a group can change because the classification changed. That does not automatically mean GA4 collected new sessions, the campaign generated more demand, or the channel became more efficient. Treat the movement as a reporting question until operational data confirms a performance change.

What to do next

  1. Save a dated comparison. Export the same reporting period using the dimensions and metrics from the prior report. Keep the original export instead of relying only on a refreshed Looker Studio dashboard.
  2. Compare dimensions side by side. Review Source, Medium, Campaign, Source Group, Source Platform, and Default Channel Group. Note which rows moved, merged, or became newly visible.
  3. Inspect tagging and integrations. Check UTMs, Google Ads auto-tagging, linked advertising accounts, cross-domain settings, consent behavior, and recent Google Tag Manager or server-side tagging changes. UTM cleanup improves future consistency but does not necessarily rewrite historical values.
  4. Audit Looker Studio. Confirm connector fields, cached data, date filters, calculated fields, blended sources, and channel filters. A dashboard change can result from connector or configuration behavior rather than campaign performance.
  5. Reconcile BigQuery by scope. Google documents separate user-, session-, and event-scoped attribution. First-user fields describe a user’s first arrival; session attribution uses the session traffic-source record; conversion-event analysis uses event-level attribution fields. Do not compare a session-scoped GA4 report directly with user- or event-scoped BigQuery analysis.
  6. Validate against business records. Compare GA4 key events with CRM leads, ecommerce orders, revenue, call records, and advertising spend. If reporting share changed while operational measures did not, treat the difference as a measurement or classification issue until further testing supports a performance explanation.

Every recurring dashboard should document its dimensions, scope, attribution model, date range, and connector. That makes GA4, Looker Studio, BigQuery, Google Ads, and Search Console comparisons easier to reconcile and reduces the chance that a reporting-taxonomy change turns into an expensive budget decision.

Sources

Need help checking this on your WordPress, Google Ads, Analytics, local SEO, or website setup? Splinternet Marketing can review the issue and help you prioritize the next fix.

This article is for informational purposes only and reflects general marketing, technology, website, and small-business guidance. Platform features, policies, search behavior, pricing, and security conditions can change. Verify current requirements with the relevant platform, provider, or professional advisor before acting. Nothing in this article should be treated as legal, tax, financial, cybersecurity, or other professional advice.

Editorial note: Splinternet Marketing articles are researched from cited platform, documentation, regulatory, and industry sources. AI may assist with drafting and review; final content is checked for source support, practical usefulness, and platform/date accuracy before publication.

Related Posts

Splinternet Marketing