- MTA:
- Multi-touch attribution is a method for assessing the effectiveness of marketing touchpoints throughout the customer journey, assigning credit to each for a more comprehensive understanding. It enables businesses to determine the influence of each touchpoint on final sales, providing a clear picture of both effective and less effective stages in the customer journey.
- Multi-touch attribution is crucial because it addresses the complexity of modern customer journeys, where customers interact with various channels before making a purchase decision.
- Notebook:
- In the notebook I begin with cleaning the dataset, analysing and implementing shapely values method to create a revenue attribution model from scratch.
- Future Work:
- Long Short-Term Memory (LSTM) networks, a type of recurrent neural network (RNN), can be employed in marketing to build sophisticated attribution models that capture the complex dynamics of customer journeys.
- The use of LSTMs will lead to superior accuracy of an attribution model because LSTMs effectively model sequential customer interactions, capturing long-term dependencies and accommodate variable sequence lengths.
- They can naturally handle time decay, where the impact of older touchpoints diminishes over time. Moreover, they allow for dynamic adjustments to the attribution model as new data becomes available. This adaptability is beneficial in marketing scenarios.