Examples of multi touch attribution rules

1 Linear
2 U-Shape or bathtub (all credited)
3 U Shape or bathtub (only 1st and last credited)
4 Linear, except for incentive channel (cashback)
5 Linear, only the last touchpoint is credited for each partner

Linear

Linear attribution means, every touchpoint gets the same credit.

The setup is quite straight forward, only the base model linear needs to be selected.

U-Shape or bathtub (all credited)

Prio on first and last touchpoints: Every touchpoint gets credited, but the first and the last touchpoint get 3x as much as the middle ones.

Base model: Linear

Create two custom rules of the type credit touchpoints

First rule: journeyPosition.fromStart = 1 and then define hoch much more credit the first touchpoint should get.

Second rule: journeyPosition.fromEnd = 1 and then define hoch much more credit the last touchpoint should get.

U Shape or bathtub (only 1st and last credited)

Only the first and last touchpoint of the user journey get (equal) credit.

Base model: Linear

Create one custom rule of the type credit touchpoints

Rule: NOT (journeyPosision.isFirst = true OR journeyPosision.isLast = true) and then define 0 times the credit.

Explanation: No touchpoint except the first and last one will get credit. The first and last touchpoint will share credit equally.

Linear, except for incentive channel (cashback)

Linear, unless there is an incentive channel (cashback) part of the customer journey.

Then, the last touchpoint of the incentive channel should get 100 % of the credit.

Base model: Linear

Create two custom rules of the type remove matching touchpoints

First rule: ujChannelIds IN (22) AND (NOT channelId = 22)

Explanation: If channel 22 appears on the customer journey, each touchpoint that is not channel 22 will be removed.

Removing a touchpoint refers only to the attribution process: it means that it cannot be credited in the same attribution model and cannot be a winner, as it will be invisible to subsequent rules.

Now, we have only touchpoints of channel 22 left in the customer journey. With the second rule we want to achieve that only the last one is credited.

Second rule: ujChannelIds IN (22) AND ( NOT(last = true) )

Explanation: If channel 22 appears on the customer journey, each touchpoint except for the last one will be removed.

The last touchpoint will be a touchpoint of channel 22, since we previously removed all other touchpoints (rule 1). Hence, the last remaining touchpoint will get all the credit.

Linear, only the last touchpoint is credited for each partner

Base model: Linear

Create one custom rule of the type remove matching touchpoints

Rule: lastSamePartnerId = false

Explanation: If a partner has multiple touchpoints in the customer journey, every touchpoint of that partner except the last one will be removed.