Fractional attribution
Fractional attribution assigns each touchpoint a fraction of a conversion rather than the whole credit, so a multi-touch path distributes one conversion across several channels. It is the mechanism behind linear, time-decay, position-based, and data-driven models, and it explains why per-channel conversion counts can be decimals that still sum to the real total.
What this means
Single-touch models give 100% of a conversion to one interaction. Fractional (multi-touch) models split that single conversion into pieces — for example linear gives each of four touches 0.25, while data-driven might give 0.4 / 0.3 / 0.2 / 0.1 based on learned patterns. The fractions for a path always sum to one whole conversion.
This is why a channel report under a fractional model can show 12.7 conversions: it holds many partial credits. The decimals are correct, not a bug — they reflect shared responsibility across the path.
Why fractions, and how to read them
Fractional attribution exists because real journeys rarely have a single cause. Splitting credit acknowledges assisting touches that single-touch models erase. The choice of model is the choice of how to split — uniformly, by recency, by position, or algorithmically.
Reading fractional reports requires care: never round per-channel fractions and then compare them to integer-counting tools; the totals reconcile but the per-channel numbers will not match a last-click report. Google documents how multi-touch models distribute credit. The headline rule is that fractions sum to conversions, and the model determines the split.
- One conversion split into fractions across touches
- Per-channel counts can be non-integer but sum to the total
- Model choice determines how the fraction is divided
How it appears in analytics and logs
Decimal conversion counts per channel are expected under fractional models; they are partial credits that add up to the integer total of real conversions.
Diagnostic use case
Use fractional attribution to reflect that several touches contributed to one conversion, accepting non-integer per-channel counts that sum to whole conversions.
What WebmasterID can help detect
WebmasterID's first-party path data lets fractional credit be reasoned about across the channels that appeared in a journey.
Common mistakes
- Treating decimal conversion counts as an error.
- Comparing fractional per-channel counts to last-click integers.
- Rounding fractions before they are summed across a path.
Privacy and accuracy notes
Fractional attribution allocates credit across recorded touches, requiring path data but not identity. This page is educational, not legal advice.
Related pages
- Multi-touch attribution: the family, not a model
Multi-touch attribution (MTA) is not one model but the whole family of models that distribute credit across more than the final touch — linear, time-decay, position-based, data-driven. What unites them is the ambition to value the full path, and the shared dependency on every relevant touch being tracked.
- Conversion credit distribution
Conversion credit distribution describes how an attribution model allocates the credit for a conversion across the interactions that preceded it. Every model — single-touch or multi-touch, rules-based or algorithmic — is fundamentally a different distribution rule. Understanding distribution as the shared concept clarifies why models disagree even on identical paths.
- Linear attribution: equal credit to every touch
Linear attribution divides a conversion's credit equally among all touchpoints in the path. It is the simplest multi-touch model: every touch matters the same. That even-handedness avoids the single-touch extremes, but it also pretends a fleeting impression and a decisive demo are worth the same — which is rarely true.
- Attribution analytics
See partial credit spread across the path.
Sources and verification notes
- Google Analytics Help — Attribution and attribution modelingDocuments multi-touch models that distribute fractional credit.
Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.