iET

iterative Expected Target is a new algorithmic attribution model that based itself on the probability of reach your target both considering the single touchpoint effect and the probability to move the users to other touchpoint. Similarly to the Markovian model attribution, iET evaluate the users transaction matrix redistributing the effect of a touchpoint to the others based on the proability to land on it. Also, it doesn't require to specify of many baksteps are necessary to consider, because of the estimation process. It could be use to evaluate a probability target (i.e. conversion rate) but also directly an economic measure like the transaction amount. The attribution values obtained with iET are in the same scale of the original target, making easy the comparison and the quantification of the impact of each touchpoint.

INSTALL

pip install iET

RESULTS