Cost Reduction and Efficiency Gains Through Migration from SAS to Open Source Solutions for a Large Media Company
Before
A large media company was heavily reliant on SAS for audience modeling and promotional campaign management. While SAS provided robust modeling capabilities, it came with significant drawbacks: high licensing and maintenance costs, data processing limits that incurred additional fees, and a limited talent pool familiar with SAS.
Solution
Our team undertook a comprehensive migration process, moving from SAS to open-source Python. We also undertook to rebuild a campaign management tool using open source frameworks, to provide a purpose built solution without prohibitive licensing fees.
Logistic Regression Models Migration: All audience modeling, previously done in SAS, was successfully transitioned to Python. These models were hosted in the client’s private cloud, ensuring security and scalability while removing dependency on costly SAS licenses.
Campaign Builder Development: We recreated the campaign builder using open-source tools, enabling the client to design and deploy promotional campaigns. The new system is lighter, easier to maintain, and fully customizable, offering the client more flexibility in managing their campaigns.
Outcome
By migrating all logistic regression models to Python and developing a custom campaign builder using open source frameworks, we delivered substantial cost savings and improved operational efficiency. The move to open source eliminated licensing and maintenance fees associated with SAS. Additionally, the company no longer faced data processing limits, allowing them to scale their campaigns without incurring extra costs. The use of Python, a more widely known language, also expanded the talent pool, making it easier to recruit and train staff.
