Features

ChannelAttribution Pro

Features available in ChannelAttribution Pro

FEATURE	DESCRIPTION	VERSION	FUNCTIONS
Unified Attribution model	Perform attribution when customer journeys are not available or when customer jouneys are available only for some channels while aggregated number of touchpoints are available for the other channels	≥ 3.8 Beta phase	uam
Transaction-level attribution with logistic regression	Perform attribution with logistic regression at transaction level	≥ 3.7	logistic_regression
Compare the performance of different attribution models	Compare the performance of last touch, first touch, linear touch, Markov model, Shapley value, and logistic regression in a traffic allocation problem based on your customer journeys	≥ 3.6 Beta phase	compare_models
Next best action with Markov model	For a partial customer journey, predict the next best action that maximizes the conversion probability of the user	≥ 3.4	next_best_action
Buget allocation with Markov model	Train a Markov model with your customer journeys and return the best budget allocation among your channels	≥ 3.3	markov_budget_allocation
Transaction-level attribution with heuristic models, Markov model and Shapley value	Perform attribution with heuristic models, Markov model and Shapley value at transaction level	≥ 3.1	heuristic_models markov_model shapley
Real-time attribution with Markov model and Shapley value	Train a Markov or Shapley model on a huge amount of customer journeys, store the model parameters and then use it to perform attribution on new customer journeys and save computational time	≥ 3.1	new_paths_attribution
Markov model and Shapley value with odds	Perform a more accurate attribution at transaction-level using odds instead of removal effects	≥ 3.1	markov_model shapley
Combine results from Media-mix Model and Multi-touch attribution models at path-level	If you have previously estimated a Media-mix model on your data, you can combine its results, at transaction level, with those of one of our multi-touch attribution model	≥ 3.1	combine_mta_mmm
Out-of-sample validation algorithm for choosing the best Markov model order	Choose the best Markov model order using an out-of-sample validation methodology, also for highly inbalanced data using precision-recall curve instead of roc curve	≥ 3.1	choose_order
Simplified Shapley value formula	Classical Shapley value formula limit the use of Shapley value to problems with less than 10 channels, while simplified Shapley value can be used also with thousands of channels	≥ 3.1	shapley
Read customer journeys directly from CSV files	Process huge amount of customer journeys avoiding out-of-memory issues	≥ 3.1	heuristic_models markov_model shapley
Multiprocessing	Faster execution of Markov model when huge amounts of customer journeys are elaborated	≥ 3.0	markov_model

ChannelAttribution Open Source

Features available in ChannelAttribution Open Source

FEATURE	VALUE	FUNCTIONS
Channel level attribution with Markov model	Split conversions and/or conversion value among your digital channels using an advanced probabilistic methodology to get a more precise measure of ROI	markov_model
Channel level attribution with Heuristic models	Split conversions and/or conversion value among your digital channels to measure ROI	heuristic_models
In-sample validation algorithm for choosing the best Markov model order	Choose the best Markov model order to perform a more precise attribution	choose_order