Glossary

Fractional attribution

Fractional attribution is a measurement approach that assigns partial credit for a conversion to several touchpoints across the entire customer journey. Rather than giving all credit to the first or last interaction, it distributes influence across every meaningful step so you can see how awareness, consideration, and decision work together.

What is fractional attribution

Fractional attribution gives each interaction along the path to conversion a share of the credit. For mobile and web journeys this can include paid and organic search, social, creator posts, email, push, in app messages, and referral. The result is a clearer view of how people moved from first interest to final action.

Example

A person sees an Instagram Story, later clicks a search ad, then converts after an email reminder. With fractional attribution, Story, search, and email each receive part of the credit in line with your chosen model.


Why it matters

Complete view of influence

You see which channels open the journey, which nurture interest, and which close the deal.

Smarter budget allocation

Spend can move from over credited last clicks to the moments that truly move people forward.

Better user experience

By understanding helpful moments across the journey, teams can remove friction and invest in content that people value.

Clearer collaboration across teams

Brand, lifecycle, and performance teams can align on shared outcomes rather than compete for single touch credit.


How fractional attribution works

  • Collect touchpoints Track impressions and clicks across channels along with key on site and in app events.
  • Choose a model Pick a rule set that assigns weights to each touchpoint.
  • Attribute conversions Distribute credit to all qualifying interactions within your lookback windows.
  • Evaluate quality Tie attributed paths to post conversion outcomes such as retention, repeat purchase, and lifetime value.

Common fractional models

Linear

Every touchpoint shares credit equally. Simple and fair when all interactions have similar importance.

Time decay

Recent interactions get more weight. Useful for fast moving decisions and short cycles.

U shaped

First touch and last touch receive a larger share. Middle touches split the remainder. Good for journeys where discovery and close matter most.

W shaped

First touch, a key middle milestone such as lead capture, and last touch receive the largest shares. Helpful for longer journeys with a distinct evaluation step.

Custom rules

You define weights that reflect your funnel. For example, raise the value of trial start or add weight for trusted creator content.

Markov chains and similar data driven methods

Instead of preset weights, these methods estimate the lift each touchpoint provides by modeling how the removal of a step changes conversion probability. This uncovers hidden helpers that rule based models can miss.


Implementation checklist

  • Map your journey and define qualifying touchpoints for impression and click based channels
  • Set lookback windows for upper, middle, and lower funnel steps
  • Use consistent campaign naming so sources can be compared over time
  • Capture identity where consented so cross device paths can be connected
  • Start with a simple model such as linear, then test U shaped or time decay
  • Validate with incrementality tests to confirm that high credited steps also drive lift
  • Report both contribution to conversions and quality outcomes such as retention and lifetime value

Pitfalls to avoid

  • Using one model as the only source of truth
  • Ignoring offline or dark social influences that are hard to track
  • Setting lookbacks that are too long which over credit old touches or too short which miss key steps
  • Letting credit drive spend without checking real lift and downstream value

Quick example

Path

Instagram Story view

Search ad click

Email click

In app purchase

Credit by model

Linear gives each step twenty five percent.

Time decay gives the email and purchase side steps the largest shares.

U shaped gives first and last the largest shares and splits the rest across the middle.

Each model answers a different question. Use two views side by side to balance discovery and closing power.


FAQs

How is fractional different from multi touch

They are the same idea. Both assign credit to several interactions rather than a single one.

Which model should I start with

Begin with linear for a baseline. Add U shaped for a brand plus performance view and time decay for closing influence.

How do I handle cross device paths

Use consented identity, logins, or publisher provided signals to link devices. Where linking is not possible, evaluate with aggregate lift tests.

Can fractional work with privacy rules

Yes. Use consented data, aggregated reporting, modeled conversions, and privacy safe APIs. Pair attribution with incrementality testing.

What metrics should I monitor beyond conversion count

Retention, repeat purchase, cohort lifetime value, and cost per incremental conversion by channel and by model.


Related Terms