Marcello Genovese on Why Strong Metrics Can Mask Broken Products
Engagement is up. Retention curves look healthy. Every dashboard metric glows green. And yet the product is fundamentally broken — a reality that won’t surface until months of runway have been spent building features nobody actually wanted. It’s a scenario product executive Marcello Genovese has watched unfold repeatedly across growth-stage companies, and his diagnosis is precise: teams confuse outputs for outcomes, and they trust intuition long after the data has stopped telling the truth.
Genovese is not describing an edge case. He’s describing a structural failure that recurs at a specific moment in a company’s lifecycle — typically after a Series A or B round, when the stakeholder base expands, the roadmap inflates, and the original user problem starts competing with investor preferences, board opinions, and competitive pressure. The product that earned early traction slowly becomes something heavier, more diffuse, and less useful. This pattern, Genovese argues, is common in growth-stage companies where decision-making becomes distributed and incentives shift.
“You see products that start strong and get traction, then they try to fulfill investor needs or CEO wishes,” Genovese says. “You should keep your vision and what you stand for, not build a product that does everything.”
His framework for avoiding that fate centers on a deceptively simple question: are we solving a problem for the consumer, or fulfilling wishes from people who don’t actually use the product? Most teams that falter during growth funding fail that test. They build for pitch meetings, internal politics, and competitive optics. The user becomes an abstraction rather than the organizing principle behind every decision.
Genovese is equally direct about the role of intuition. Founders who achieve early traction frequently develop high confidence in their own judgment. Positive numbers reinforce that confidence. What doesn’t show up in aggregate statistics — the friction, confusion, and indifference that real users quietly experience — goes unexamined. Intuition, in this context, becomes a liability.
“People misunderstand that nothing is as easy as it looks,” he observes. “You still need to think through your product, and a good product strategy is always about the end user and how they’re going to use it.”
The corrective Genovese advocates is not bureaucratic process but disciplined validation — specifically, rapid prototyping with minimal investment before significant engineering resources are committed. He is explicit that polish is irrelevant at this stage. “I test products with simple prototypes that may look ugly, but they have the functionality, and that helps.” Rough prototypes expose flawed assumptions early. If a stripped-down version cannot demonstrate value, a polished one will reach the same conclusion at greater expense.
When teams have already traveled far down the wrong path, Genovese offers advice that is uncomfortable but, in his assessment, often necessary: throw it out and rebuild. “Be bold enough to throw away what you’ve done and start from scratch,” he says. “I’ve seen products improve their design and structure when they started from scratch and rethought from the beginning what problem they’re actually solving.” Sunk cost psychology is particularly severe at the growth stage, where months of work and millions of dollars make any pivot feel like failure. Genovese frames the decision around the original vision — whether the product still solves the problem it was built to address, or whether it has drifted into something that serves internal stakeholders more than users.
He also flags a paradox introduced by modern tooling. AI and improved development infrastructure have made building faster and cheaper, which removes the natural friction that once forced teams to validate before committing. Speed now enables teams to travel farther down the wrong path before recognizing the mistake. The appropriate response is to use that speed for learning — testing more hypotheses cheaply — rather than shipping more features without evidence that users want them.
“Talk to your users, do reviews, run user testing,” Genovese says. “There are plenty of testing platforms out there.”
The throughline across Genovese’s thinking is that process, correctly understood, is not a constraint on speed or creativity. It is the mechanism by which teams stay honest about who they are building for. Intuition is useful when it has been trained by consistent contact with real user behavior. When it hasn’t — when it is based on internal conviction, board expectations, or wishful interpretation of metrics — process is what keeps a company from spending its way into a product nobody needs.