Author(s): Jonty Haberfield

Originally published on Towards AI.

A tour of three different MTA approaches — Shapley, Markov, and LSTM — covering theory and practice

An email. A push notification. An offer in a loyalty app. A Google ad. A banner on a website. Another email. Finally, you relent and buy. But the marketing team wants to know — why did you buy? How much did each of those touchpoints contribute to your eventual decision?

This article explores various multi-touch attribution (MTA) methods, including Marketing Mix Modeling, Incrementality Testing, and multi-touch attribution dominated by user-level metrics. Each method weighs the contributions of different marketing channels in driving conversions, with the discussion highlighting the complexities involved in different attribution models. It evaluates empirical approaches, such as Shapley values and Markov chains, and contrasts traditional attribution strategies with advanced techniques like LSTM that incorporate deep learning, illustrating the significant discrepancies in conversion crediting across methodologies while emphasizing the necessity of choosing appropriate models based on specific marketing needs.

Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI