The Double-Edged Sword: Data vs. Opinion in Product Management
"Without data, you are just another person with an opinion." We’ve all heard it, and there’s a lot of truth to it. Data is indispensable for understanding user behavior, measuring impact, and making informed decisions. But I’ve also come to firmly believe the corollary: "Without opinion, you are just another person with data."
In product management, finding the right balance between data-informed insights and strong, well-reasoned product opinions is a constant dance. Neither can stand alone if you want to build truly great, innovative products.
When Data Shines (And When It Can Mislead)
Data is fantastic for:
Identifying Trends and Patterns: Understanding what users are doing.
Measuring Progress: Tracking KPIs and the success of initiatives.
Validating Hypotheses: Running A/B tests or experiments to confirm assumptions.
Spotting Anomalies: Noticing shifts in metrics, like a sudden drop in NPS contributors (which can be a bigger red flag than the NPS score itself!), can signal underlying issues or successes.
However, data alone can sometimes be misleading or insufficient:
Local Optima: Over-reliance on A/B testing incremental changes can lead you to a local maximum, missing out on bigger, bolder innovations.
The "Why" Gap: Data tells you what is happening, but often not why. Qualitative insights are needed to understand the motivations and frustrations behind the numbers.
Analysis Paralysis: Drowning in data without a clear hypothesis or product sense can lead to inaction.
Misleading Metrics: As I’ve often said, be wary of cumulative numbers for customers or revenue. They almost always look like a hockey stick, even if the underlying health of the business is poor. Focus on cohort behavior and true growth rates.
The Power of Product Opinion (When It's Not Just Guesswork)
A strong product opinion isn't about gut feelings in a vacuum. It's a perspective shaped by:
Deep Customer Empathy: Truly understanding user needs, pain points, and desires.
Market Understanding: Knowing the competitive landscape and broader trends.
Technical Feasibility: Having a grasp of what's possible.
Vision and Strategy: A clear idea of where you want the product to go.
Common Sense & First Principles: Sometimes, if you strongly believe something should work based on fundamental user needs, it probably does, even if initial data is ambiguous.
Great PMs use data to refine their opinions and test their hypotheses, but they also use their opinions to guide where they look for data and what problems they choose to solve.
Avoiding Self-Fulfilling Prophecies
One trap I always caution PMs against is using data to validate a mediocre idea simply because it’s their idea. If you’re testing a prototype that even you feel has a clunky UX, don’t waste time on endless customer interviews to confirm what you already suspect. Only invest heavily in testing solutions that you and the team genuinely believe are outstanding.
The most effective product leaders and teams I know don't see data and opinion as opposing forces. They see them as complementary tools, both essential for navigating the complexities of building products that customers love and that drive the business forward. It's about being data-informed, not data-dictated, and having the courage of your (well-reasoned) convictions.
