In a career that spans decades helping build products for some of the most successful companies in the world, including Hewlett-Packard, Netscape Communications, America Online and eBay, Marty Cagan knows what it takes to innovate – and what gets in the way.

Cagan, founder of the Silicon Valley Product Group, spoke at Intuit’s recent Create the Offering Forum, an annual event dedicated to inspiring product developers, managers and designers to create awesome products.

In his wide-ranging presentation, Cagan covered topics ranging from customer insights to analytics, and offered advice that applies to companies trying to innovate to deliver awesome products. Here is some of what he shared.

Leadership over management. “At middle management, teams desperately need true leadership,” he said. Leaders need vision. “Not a little vision, but the whole ecosystem you’re trying to create. Someone has to carry that torch.”

Know what you can’t know. “Business plans are OK,” he said, but can be ineffective for long-term product development. “You can’t know which ideas your customers are going to respond to and use. The good teams count on at least three-fourths of their ideas not working.”

Know what your customers can’t know. Cagan urges talking to customers, with a caveat. “When you go to a home or small business, you’re not there to ask them what they want. Customers don’t know what’s possible. It’s your job to know what’s possible. I’m not suggesting that you stop talking to customers. You should test your ideas on customers.”

Discovery over roadmaps. Quarterly and annual roadmaps can inhibit discovery, Cagan said. “If you’ve got a year committed to new features, you’re not committed to discovery. You’re going through the motions.” Cagan solidly endorsed the idea of rapid iteration, urging leaders to test products within two weeks. “Never wait more than two weeks before you’re testing your idea on real customers. If you go longer than two weeks, it’s your baby. You’re too attached.”Cagan_small

Data over opinions: “You need to get to the point where you’re not arguing over everything. You just run a test,” he said. “You use the data to drive a decision…otherwise you’re flying blind. I don’t put a feature in if we don’t have analytics to see if it actually works.”

Outcome over output: “Focus on outcome, not output,” he said. “You can be a crummy team using very good processes. That’s not what it’s about. It’s about results.”

Collaboration over consensus; small co-located durable teams. “It is not about consensus. (When) everybody’s trying to accommodate everybody, it causes the wrong results and the wrong response,” he said.

In addition, Cagan endorsed collaboration among small, co-located teams. “Co-location means a product manager, product designer and engineers are this close. We’re going to lunch together, hanging out together. We don’t need meetings. We look at the screen and see. It’s a different dynamic. Co-located teams are better than distributed teams.”

Culture over process. Regimented process can undermine innovation, Cagan said. “You cannot have that mindset. It’s not about process. It’s about culture. (Process) creates a mindset, a bureaucracy. The gauntlet you have to run is crazy.”

Photo: Intuit founder Scott Cook (left) and Marty Cagan on stage at Intuit’s annual Create the Offering Forum.