Attribution

Attribution ties a conversion or outcome back to the specific point, decision, or campaign that influenced it. Without it, you know personalization ran — not whether it actually moved the result you care about.

What attribution connects

What an attribution result ties together
Element	What it tells you
Conversion/outcome	The event you defined as success — a purchase, signup, or other goal.
Decision	Which specific decision served the content or recommendation involved.
Point	Which personalization point the decision ran in.
Campaign	Which campaign the point and decision belong to, if they're grouped under one.

Attribution models, generally

Attribution logic generally falls into a few categories: single-touch (all credit to one touchpoint — the first or the last), fractional/multi-touch (credit split across several touchpoints a visitor encountered before converting), and algorithmic models that weight touchpoints statistically rather than by a fixed rule.

Read an attribution result

Using attribution results

Start from the conversion you defined, not the decision — work backward to see what influenced it
Check which point and decision are credited before assuming a specific campaign drove the result
Compare results across decisions in the same point to see which is actually converting better
If a decision or point you expect to see is missing from results, confirm it's enabled and has run recently — see Personalization logs

For what each decision run itself recorded (content served, timing, outcome), see Personalization logs. For the underlying data model attribution and logging are built on, see the DMO reference.

Discussion

Comments are not configured for local development. They appear once the site owner connects GitHub Discussions (Giscus).