Marketing Channel Performance in GA4: Tracking New vs. Returning Users

Understanding how different channels contribute to user acquisition vs. user retention is critical for optimizing spend and strategy. In GA4, while you get some visibility into user types, a more detailed analysis often requires BigQuery SQL.

This guide walks you through:

1. Why this analysis matters
2. How GA4 tracks user types
3. Step-by-step: Build a New vs. Returning User breakdown by marketing channel using BigQuery
4. Key insights to act on

Why This Analysis Matters

Most marketing teams ask:

• Are we just getting clicks, or are people coming back?
• Which campaigns drive long-term value, not just first visits?
• Is email working for re-engagement?
• Do paid ads convert once, or create loyal users?

By breaking down New vs. Returning Users by Channel, you can double down on channels that drive loyalty, not just top-funnel awareness.

How GA4 Tracks New and Returning Users

GA4 Documentation: Measure New and Returning Users

How to Do It in BigQuery

Step 1: Start with GA4 BigQuery Export

We’ll use the events_YYYYMMDD table that includes all session data and event-level parameters.

WITH base AS (
  SELECT
    user_pseudo_id,
    MIN(event_timestamp) AS first_seen,
    user_first_touch_timestamp,
    session_source,
    session_medium,
    session_campaign,
    event_date
  FROM
    `your_project.analytics_XXXX.events_*`
  WHERE
    _TABLE_SUFFIX BETWEEN '20250801' AND '20250806'
    AND event_name = 'session_start'
  GROUP BY
    user_pseudo_id,
    user_first_touch_timestamp,
    session_source,
    session_medium,
    session_campaign,
    event_date
),

classified AS (
  SELECT
    *,
    CASE
      WHEN event_timestamp = user_first_touch_timestamp THEN 'New'
      WHEN event_timestamp > user_first_touch_timestamp THEN 'Returning'
      ELSE 'Unknown'
    END AS user_type
  FROM base
)

SELECT
  session_source,
  session_medium,
  user_type,
  COUNT(DISTINCT user_pseudo_id) AS users
FROM classified
GROUP BY 1, 2, 3
ORDER BY users DESC;

Example Output

Interpretation:

• Google Ads (CPC) brings in lots of new users but has lower return rates.
• Email, while driving fewer new users, is excellent for bringing people back.

Key Considerations

• Attribution Window: This SQL uses first-touch attribution; customize if needed.
• Session Accuracy: Check how session_source is populated in your setup.
• Cross-device Tracking: If you’re not using User-ID, returning users on a new device may appear “new.”

Tips for Marketers

• Prioritize retention-focused campaigns (email, push) if LTV is key.
• Use GA4 Audiences to retarget Returning Users with different messaging.
• Test channel-specific onboarding flows — what makes a new user return?