Skip to content

Stockertim Dictionary · Marketing

Attribution model

Also called marketing attribution, multi-touch attribution, attribution modelling

In short

A rule for sharing the credit for a sale or other conversion among the marketing touchpoints that came before it.

  • Intermediate
  • Concept
  • Checked

In detail

Most purchases follow several contacts with a brand: a search advert, a social post, an email, a banner, a direct visit. Each channel's own report tends to claim the whole sale, so five reports can describe five sales where there was one. An attribution model replaces those competing claims with one stated rule for dividing the credit, applied to every conversion so that channels can be compared.

Common rules include last click, which gives everything to the final touch; first click, which rewards the touch that introduced the buyer; linear, which shares credit equally; time decay, which gives more to touches nearer the sale; and position-based, which gives large shares to the first and last touches and divides the rest. Data-driven models compare paths that ended in a sale with paths that did not and estimate each touch's share statistically.

Every rule embeds an opinion about how marketing works, and the choice can move a channel from looking like the best buy to looking like a waste. Last click flatters channels that close, such as brand search; first click flatters channels that find new people. Teams therefore choose a model to suit the decision in hand and record why.

All models share a blind spot: they can share credit only among touches that are tracked. Television, posters, word of mouth and journeys across devices or into a shop leave gaps. A model also describes sequence, not cause, so channels whose credit swings most between models are candidates for an experiment.

Formula

Linear credit per touch = conversion value / number of touches

conversion value
the revenue or count of the conversion being shared
number of touches
the tracked contacts before the conversion

Worked example

A £120 order follows four tracked touches: a search advert, a social post, an email and a direct visit.

  1. Linear share for each touch: 120 / 4 = £30
  2. Position-based share for the first touch: 120 × 40% = £48
  3. Position-based share for each middle touch: 120 × 10% = £12

Last click would give the direct visit all £120, while the linear rule gives it £30: the same order looks very different under each rule.

In practice

Analytics tools and advertising platforms offer several models. Marketing teams pick one for channel reporting, compare it with a second to see which results depend on the rule, and test doubtful channels with control groups.

Watch out

  • Attribution shows which touches came before a sale, not which ones caused it.
  • Adding up each platform's own credited sales overstates the total.

Learn it properly

Official sources

Facts last checked on 3 November 2026.

Education, not advice: the Stockertim Dictionary explains what terms mean. It never recommends buying or selling anything.

English