Data-Driven Attribution in GA4: The Three Channel Reports Explained

Data-Driven Attribution in GA4: The Three Channel Reports Explained

Data-driven attribution (DDA) in Google Analytics 4 distributes conversion credit across all touchpoints in the user journey based on actual interaction patterns, rather than assigning 100% credit to the last click. It is the default model for GA4's Advertising reports, but not for the Traffic acquisition or User acquisition reports, which still use last-click. This single difference causes most cross-channel budget decisions made from GA4 to be based on the wrong attribution model.

Key takeaways

1. GA4 has three separate "channel group" parameters: Default channel group (event scope, data-driven attribution), Session default channel group (session scope, last-click), and First user default channel group (user scope, last-click). Each returns different conversion numbers for the same channel.
2. Comparing "Paid Social" in the Traffic acquisition report to "Paid Social" in the All channels report is a systematic attribution error, the two reports use different models and will never agree.
3. Data-driven attribution requires a minimum of 300 conversions per 28-day period per event to train the model. Below this threshold, GA4 silently falls back to last-click even when DDA is labeled in property settings.
4. The correct GA4 parameter for budget allocation decisions is the Default channel group under event scope with DDA (Advertising section), not the session-scoped parameter in Traffic acquisition that most teams use by default.
5. Prooflytics standardizes which GA4 parameter and attribution model backs every number in the daily briefing, preventing the silent attribution mismatch that causes budget decisions to be made from the wrong report.

Why GA4 attribution confuses performance teams

The ICP problem this creates for in-house analytics teams: two people pull "Paid Social conversions last month" from two different GA4 reports and get different numbers. One says 400. One says 650. Neither is wrong. They are looking at different attribution models applied to different scopes. The budget decision gets made from one of them, and the team has no way to know which is the more accurate picture of channel contribution.

Attribution model: a rule that determines how conversion credit is distributed across the marketing touchpoints in a user's path to conversion. Last-click gives 100% credit to the final touchpoint. Data-driven distributes credit based on the actual measured contribution of each touchpoint using a machine learning model.

Attribution scope: the level at which credit is assigned, event (this specific conversion), session (the session where the conversion happened), or user (where the user was originally acquired).

The three GA4 channel group parameters

GA4 exposes three distinct channel group parameters. They are not variants of the same metric. They answer different questions and use different attribution models.

01. Default channel group (Event scope, Data-Driven Attribution)

Found in: Advertising section, "All channels" report and Conversions report.

This parameter shows which channel received DDA credit for the conversion event. Because DDA distributes fractional credit across multiple touchpoints, the sum of conversions across all channels can exceed total actual conversions. Each channel receives a fractional credit that sums to 1.0 per conversion.

Use when: the question is "which channel contributed to this conversion?" This is the most accurate model for cross-channel budget decisions.

02. Session default channel group (Session scope, Last-Click paid + organic)

Found in: Acquisition section, "Traffic acquisition" report.

This parameter credits the channel that started the session in which the conversion occurred. Attribution model is last-click, including both paid and organic channels. A conversion is attributed to the session's source channel regardless of which paid ad was seen earlier in the user journey.

Use when: the question is "what session contained the conversion?", primarily for session-level behavior analysis, not for channel ROI allocation.

03. First user default channel group (User scope, Last-Click paid + organic)

Found in: Acquisition section, "User acquisition" report.

This parameter shows where the user first came from when they visited the site for the first time ever, regardless of what channel was active at conversion time.

Use when: the question is "where do our acquired users originally come from?", useful for new user acquisition strategy, not for in-month performance optimization.

Prooflytics

Turn scattered analytics into one clear picture

Every source in one brief. The whole picture. Your decision.

Start free trialSee pricing

14 days free · no credit card

What the data shows: the GA4 attribution parameter mismatch

The ICP problem this creates for in-house analytics teams: the default instinct is to pull the Traffic acquisition report because it shows a familiar sessions-and-conversions-by-channel table. But that report uses session-scoped last-click attribution. The result: Paid Social and TikTok appear to underperform (because they contribute at the discovery stage, not the final session before conversion), and Brand Search appears to overperform (because brand searches happen close to conversion). The budget shifts toward Brand Search and away from discovery channels, which eventually kills the top-of-funnel driving those brand searches.

The GA4 THREE CHANNEL GROUP PARAMETERS framework, documented in GA4's official measurement specification, names this mismatch structurally. The antipattern: comparing "Paid Social" from the Traffic acquisition report with "Paid Social" from the All channels report. Different attribution models, different scopes. The All channels DDA report will show materially higher Paid Social conversions because DDA credits touchpoints earlier in the conversion path.

The decision rule from the GA4 framework:

• "Which channel produced this conversion?", use Default channel group, event scope, DDA (Advertising section)
• "Which session contains the conversion?", use Session default channel group (Traffic acquisition)
• "Where did our users come from originally?", use First user default channel group (User acquisition)

The operational implication: most teams are making channel budget decisions from the Traffic acquisition report (session-scoped, last-click) when the DDA event-scoped report in the Advertising section gives a more accurate picture of channel contribution. Switching to the correct report typically increases credited conversions for upper-funnel channels by 20-40%.

Prooflytics standardizes which GA4 parameter and attribution model backs every number it surfaces, preventing attribution mismatch from appearing as a channel performance problem in the daily briefing.

How data-driven attribution actually works

DDA in GA4 uses a machine learning model that analyzes the paths of users who converted and users who did not. It measures the counterfactual: what would have happened to the conversion probability if a specific touchpoint had been absent from the path?

Three requirements for DDA to be active:

1. 300+ conversions per 28-day period for the specific conversion event. Below this, GA4 falls back to last-click for that event automatically.
2. 600+ non-conversion paths in the same period, the model needs both converted and unconverted paths to measure incremental contribution.
3. The GA4 property linked to Google Ads if DDA is being used for Smart Bidding, the signal flows from GA4 to Google Ads via the linked property.

If GA4 settings display "Data-Driven" as the model but the property has fewer than 300 monthly conversions for a specific event, DDA is labeled but last-click is applied. Verify at the event level in Admin > Attribution settings > "Reporting attribution model", it shows per-event status.

GA4 has 18 default channels that cannot be edited

GA4 defines exactly 18 default channels. They are not configurable. Custom channel groups can be created in GA4, but the default channel group that feeds the DDA model uses these 18 categories:

Direct, Cross-network, Paid Shopping, Paid Search, Paid Social, Paid Video, Paid Other, Display, Organic Shopping, Organic Social, Organic Video, Organic Search, Email, Affiliates, Referral, Audio, SMS, Mobile Push Notifications.

Knowing this list matters because traffic that does not fit these categories falls into Direct or Unassigned. If a significant portion of your paid traffic appears as Direct in GA4, UTM parameter coverage is the likely cause, not attribution model behavior.

Common attribution errors and how to fix them

Error: using Traffic acquisition for budget decisions This report uses session-scoped last-click. It will always under-credit discovery channels (Paid Social, TikTok, Display) and over-credit Brand Search. Fix: use Advertising > All channels for cross-channel budget decisions.

Error: assuming GA4 and your ad platform agree on conversions They will not agree. GA4 uses DDA across all channels; Meta uses its own last-click model (7-day click, 1-day view); Google Ads uses Google-attributed DDA or last-click at the keyword level. Budget decisions should be made from GA4's DDA model as the neutral third party.

Error: not verifying DDA is actually active per event GA4 shows DDA as the property-level setting, but individual conversion events revert to last-click if they fall below the 300-conversion threshold. Always verify at the event level, not just the property level.

Bottom line

• GA4's three channel group parameters return different numbers for the same channel because they use different attribution models and different scopes, using the wrong one for budget decisions is a systematic error.
• Data-driven attribution (event scope, Advertising > All channels) is the most accurate model for cross-channel budget decisions because it distributes credit based on measured touchpoint contribution.
• DDA requires 300+ monthly conversions per event to activate; small accounts see last-click applied even when DDA is displayed in property settings.
• The most common error: cutting discovery channels based on Traffic acquisition report data (session-scoped, last-click) that structurally under-credits upper-funnel touchpoints.
• You can read independent reviews of Prooflytics on G2 and compare it to alternatives in the marketing analytics category.

Start the 14-day trial, no credit card required.