United StatesHealthcare and life sciences2018 – ongoingCVS Health
Machine learning personalization for pharmacy customers
CVS Health raised medication adherence 1.6% with machine learning that tailors pharmacy outreach
Problem
CVS Health runs about 10,000 US stores and wanted to personalize outreach to pharmacy customers. Its Hadoop environment hit hardware and storage limits. A first personalized campaign reached 1% of customers and could not scale to 5%.
Approach
CVS moved the work to Azure Databricks. Machine learning models predict the best channel and timing for prescription renewal offers and side-effect check-ins. They also estimate how likely a customer is to buy a given product.
Results
Takeaway
The personalization models were limited first by infrastructure. Moving platforms let CVS scale a campaign it had stalled at 1% of customers.
Categories
- Size
- Large enterprise
- Industry
- Healthcare and life sciences
- Business function
- Marketing and sales
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks describes how CVS Health moved from Hadoop to Azure Databricks to scale machine learning personalization for pharmacy customers. The models choose communication channels and timing for renewal offers and check-ins. The page reports a 1.6% increase in medication adherence.
Full details are in the original sources.