BigQuery Attribution Models

Case study

The purpose of this project is to assist clients in understanding the attribution channels that are most influential in prescribing their drug, by doctors. By analysing data from various touchpoints/marketing channels and combining it with healthcare professional and claims data, the goal is to identify the key drivers behind prescriptions and their impact on increasing sales.

The Challenge

The client has six touchpoints/marketing channels through which their marketing/sales team reaches out to the clients (HCPs and HCOs) – Call Activity, VAE Activity, Speakers, SFMC Emails, Website and PLD to educate the clients about their drug.

Tech Mahindra team in collaboration with Google designed the solution to give the attributions of channels which are contributing to prescribe OR create a claim.

The model identifies which channels are most influential in prescribing the drug by doctors – which in turn impact the increase in sales.

The Solution

We supported The client with the solution over the current process using Google Services such as Artificial Intelligence, Machine Learning, and Data Analytics.

The solution is powered by a combination of rules & data-driven multi-touch attribution model. The AI/ML models are built using GCP Big Query with the available algorithms. The current Veeva landscape data is available in GCS bucket.

Our Approach

The approach for achieving these objectives involves:

Data gathering and preprocessing

Feature engineering and model development

Model training and evaluation

Interpretation of model results and feature importance analysis

The Challenge

Consumer misinterpretation of CBD products impacting sales in both ‘bricks & clicks’

Limited marketing opportunities due to regulatory restrictions to help reframe and alleviate concerns.