Leveraging Data Science

Cagan argues that many product companies have yet to seriously consider data science, even though leading firms are already gaining advantages from it. He distinguishes data science from traditional analytics by its focus on predictive, forward-looking insights made possible by techniques like clustering, regression, and machine learning combined with big data tools. He identifies two main applications: using data science internally to tune the product and business, and embedding it into the product to create customer value. The essay closes with practical guidance on sequencing, hiring, and organizing data science capability so the expertise spreads across the company rather than staying siloed.

01Key takeaways

• Treat data science as a source of predictive insight, distinct from backward-looking analytics.
• Pursue internal-insight uses first if you lack data infrastructure, since basic analytics often delivers critical value.
• For customer-facing opportunities, bring in expertise quickly to assess what is possible before competitors move.
• Hire data scientists who care about the product or business problem, not just the math.
• Separate the data scientist role from the data infrastructure engineer role, as they serve different needs.
• Build broad 'data IQ' across the product organization rather than isolating data science in a service team.

02Key sections

What data science adds

Data science uses statistical techniques on collected data to uncover new, often predictive insights that go beyond traditional reporting about the past.

Internal insights

Teams apply techniques like clustering and regression to segment users and tailor experiences or messaging to drive desired behaviors, such as conversion in a freemium model.

Customer value

Data science becomes part of the shipped product itself, such as personalized recommendations or machine-learning-based spam filtering.

Building the capability

Establish basic data infrastructure and analytics before pursuing data science, and figure out where data scientists sit organizationally.

Hiring and roles

Prioritize candidates passionate about the business problem, distinguish data scientists from data infrastructure engineers, and recognize that other roles can learn these skills.

Avoid siloing

Spread data thinking across the organization through all-hands, write-ups, and embedded team members instead of treating data science as a separate service.

03From the post

“A partnership dedicated to teaching best practices to product teams and product leaders”

04Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.