The $5M Netflix Mistake Most Data Analysts Still Make
Netflix faced a roughly $5 million mystery in broken DVDs. Customers looked responsible. They weren't. One unexpected data connection turned a misleading correlation into an actionable supplier decision.
A roughly **$5 million problem with damaged DVDs** at Netflix seemed to point toward customer behavior and movie genres. My first hypothesis was wrong. Following the evidence eventually led us to a DVD manufacturer. The bigger lesson isn't about DVDs. It's about **choosing the question worth answering** before diving
Read full description
into the data: - **Who** can make a decision based on the answer? - **Which questions** can that person act on? - **What is each answer worth** to the business? - **What can we realistically answer** with the time and evidence available? Rank those questions, test the strongest hypothesis, and change direction when the evidence disagrees. Correlation can provide a useful lead without explaining the cause. **The deliverable wasn't the analysis. It was a better decision.** *Originally published on YouTube September 4, 2026.*
Keywords
- Netflix DVD breakage
- root cause analysis
- correlation vs causation
- business impact of analytics
- data-driven decision making