Google Analytics 4 added a Source Group field in June 2026: a new dimension that consolidates the scattered source values a single platform can generate — facebook.com, m.facebook.com, l.facebook.com, instagram, fb, and similar variants — into one standardised bucket per platform. For teams that have spent years hand-building regex-based channel groupings to catch these variants, it’s a genuine time-saver. For teams that haven’t audited what their existing custom channel groups actually catch, it’s a quiet trap: GA4’s new default grouping and your old custom one can now disagree about which sessions belong to which platform, and both will keep reporting numbers with total confidence.
The problem shows up as a step change in a channel’s reported volume the moment Source Group rolls into a report, with no change in actual traffic. A social channel that looked stable for months can suddenly jump or drop as sessions get reclassified under the new grouping logic, and if nobody’s watching for the transition, that shift gets read as a real trend — a campaign working better, a platform sending less traffic — when it’s actually a measurement definition changing underneath a chart that looks the same as it did last week.
Data Points to Track
- Session count by Source Group versus your existing custom channel grouping, for the same date range, to quantify exactly how much reclassification the new field introduces before trusting it in a report
- Source value diversity within each Source Group bucket, so you know how many distinct raw source strings (
facebook.com,fb,m.facebook.com) are actually being folded together for each platform - Before/after conversion rate per channel, checked across the Source Group transition date, to confirm a rate change is real and not an artefact of sessions moving between buckets
- Unclassified or “other” session share, tracked over time, since a rising share here means new source variants are appearing that neither GA4’s default grouping nor your custom rules currently catch
- Cross-report consistency, comparing any dashboard still built on the old custom channel grouping against ones that have switched to Source Group, so stakeholders aren’t shown two different “Facebook traffic” numbers without knowing why
Setup Steps
- Pull a session-count comparison between Source Group and your existing channel grouping for a recent 30-day window before changing any dashboard, so the delta is measured rather than guessed.
- Audit your custom channel grouping rules for source variants that Source Group now handles natively, and note any variant your rules previously caught that the new field appears to miss.
- Flag the exact date Source Group data starts appearing in your property, and annotate any dashboard spanning that date so a step change is understood as a definitional shift, not a traffic shift.
- Decide on one canonical grouping — Source Group or your existing custom rules — for external reporting, rather than letting different teams pull from different definitions of “social channel.”
- Re-run any historical channel-performance comparison (this quarter vs. last) using a consistent grouping method across the whole period, since mixing definitions mid-comparison will misstate the trend either direction.
Actionable Insights
A large gap between Source Group and your old custom grouping for a specific platform means your legacy regex rules were either overcounting or undercounting that channel for as long as they’ve been in place — worth establishing which one before picking a canonical source of truth going forward. A rising “unclassified” share after adopting Source Group is a sign new source-string variants are emerging faster than either grouping method accounts for, usually from a new referral integration or an ad platform change, and is worth chasing down directly rather than letting it accumulate as noise. And any conversion-rate swing that lines up exactly with the Source Group rollout date, with no corresponding change in spend or campaign activity, should be treated as a reporting artefact first — check the reclassification before reporting a real performance change upward.
Related Resources
Need help tracking this in your app?
Our team sets up analytics pipelines for mobile and web teams every day. Talk to us and get your first events flowing in under an hour.
Talk to an expert