Paul Smolikov

About Paul Smolikov

20 years’ experience in data science supporting senior leadership across Finance…

Experience

1. Senior Director, Data Science

   Microsoft

   Oct 2001 — Present · Redmond, WA, US

   Supporting Finance, Business Planning, Strategy, Surface and Windows Marketing teams for 900M+ users world-wide in the following areas:Acquisition Modeling: • Classification and deep learning models for prospective buyers of all 1st party devices.• OEM Devices end-of-life models (Premium, Core, Budget and Gaming OEM PCs).• Windows 10 and Windows 10 S upgrade models.Retention/Engagement:• Classification models for Surface comparable (premium 2-in-1 OEM) and competitors’ (iPad Pro, MacBook, Chromebook) devices. • Predictive churn models for 1st party devices and Windows 10.• Retention drivers’ models. Customers’ Life-Time-Value Modeling:• User level incremental LTV models to understand the effect device purchase or service adoption on customer’s monetization level.• Revenue attribution models to understand network effect (affinity) between different services.• Customer Future Value models to isolate people we want to break up with vs. people with significant future potential.Strategy Optimization and Experimentation: • Marketing Optimization Framework using Monte-Carlo simulations to estimate monetary return of different decisions based on expected LTV and propensity to adopt a service/buy a device.• Persona management Custom Audiences Framework.• New Services Adoption Paths (Markov chain model) to predict most likely transition path• Microsoft Ecosystem Health Index to measure cross-service usage within Microsoft ecosystem. • Search User Segmentation to understand search user’s intent.