Attribution Models
Attribution defines how conversions (sales) are credited to ad interactions.
In GoWit, attribution works along two main dimensions:
-
Interaction Type
- Post-Click Attribution → A conversion is counted only if the user clicked the ad before the purchase.
- Post-View Attribution → A conversion can be counted if the user only viewed the ad (impression), even without clicking, as long as the purchase happens within the attribution window.
Example:
- Post-Click: User clicks ad for Product A → Purchases Product A.
- Post-View: User sees ad for Product A (no click) → Purchases Product A.
-
Sale Type
- Direct Sale → User purchases the exact product shown in the ad.
- Indirect Sale → User purchases a different product from the same advertiser (halo effect).
Example:
- Direct: Ad shows Product X (Brand Y) → User buys Product X.
- Indirect: Ad shows Product X (Brand Y) → User buys Product Z (Brand Y).
Combined Attribution Paths
By combining both dimensions, GoWit supports 4 attribution paths:
- Post-Click Direct – Click + same product.
- Post-Click Indirect – Click + different product (same advertiser).
- Post-View Direct – View only + same product.
- Post-View Indirect – View only + different product (same advertiser).
Attribution Window
- Attribution is evaluated within a configurable conversion_range (lookback window).
- This window defines how far back from the sale event GoWit will check for ad interactions (clicks or views).
- Both post-click and post-view events use the same conversion_range setting — the window does not differ by event type.
- Events that fall outside the defined window (e.g., impressions or clicks too old relative to the purchase) are not considered for attribution.
Attribution Logic in Reporting
- Both Post-Click and Post-View conversions are always calculated internally.
Attribution Settings in Admin Panel
There are two levels of attribution settings available in the Admin Panel:
-
Attribution Settings by Format
- Location: Admin Panel → Settings → Campaign Settings → Conversions
- Defines which attribution type (Post-Click or Post-View) is applied by default to each campaign format (e.g., Sponsored Product, Sponsored Display, Sponsored Video).
- This determines the attribution type shown in dashboard tables.
- Example: If Sponsored Product is set to Post-Click, all conversion metrics in tables will reflect Post-Click attribution.
-
Attribution Settings by Reporting
- Location: same Conversions tab, under Reporting section.
- Enables both Post-Click and Post-View metrics to appear side by side in reports.
- Options:
- Separate Metrics in Marketplace Report → Splits attribution metrics in marketplace-level reports.
- Separate Metrics in Advertiser Report → Splits attribution metrics in advertiser-level reports.
- When enabled, both metrics are visible, helping compare click-driven vs. view-driven conversions.
Important Note for Testing
- Attribution configuration is primarily a business decision, but it also affects what results are visible during testing.
- If you are testing on a format configured with Post-Click attribution only, you may not see view-through conversions even though they are tracked.
- To validate full integration during testing, ensure you understand which attribution mode is active in the settings.
- Enabling Separate Metrics during QA can be helpful to confirm that both Post-Click and Post-View events are flowing correctly before launch.
⚠️ Best Practice:
- Keep attribution settings consistent during a campaign’s lifetime for reliable ROAS/ACOS tracking.
- If you want to see both models at once, simply enable the separate metrics option in the Conversions settings (no need for custom reporting request).
Integration Requirements
For attribution to function correctly:
- Use stable
session_id
across impressions, clicks, and sales. - Include customer_id if available (for cross-device attribution).
- Always send all sales (including organic). GoWit will filter attribution automatically.
- Ensure timely event logging (impressions, clicks, sales).
Why It Matters
Attribution models affect:
- Which conversions are counted for reporting.
- How advertisers evaluate campaign performance.
- Metrics like ROAS, ACOS, CVR.
Correct setup ensures advertisers see an accurate and consistent view of ad impact.