Steven Sinofsky and Suhail Doshi - Mixpanel
Blog Post

Steven Sinofsky and Suhail Doshi

Customer Advocacy @ Mixpanel

Last night, Andreessen Horowitz board partner and former president of Microsoft’s Windows division Steven Sinofsky joined Mixpanel CEO Suhail Doshi  at Mixpanel’s Office Hours to discuss the renaissance going on in product right now. 

Suhail and Steven discussed which skills have remained essential to product management, how the industries of old should adapt to product management, and how product managers should adapt to the established institutions.

Here is a list of some key insights from the evening. (Of course, it’s better to get it straight from the experts, so we’re re-broadcasting the whole talk on Thursday, February 2—sign up to watch here.)


1. Software is still eating the world

Steven started the evening by echoing the words of a16z founding member Marc Andreessen: Software is eating the world. Since the dawn of industry, people have been building products, whether they’ve been called that or not. But with that newer distinction, product managers have come to the forefront—and data is coming along with them. Steven said:

“Everything that gets made has some sort of notion of a product manager, even if they didn’t always call it that. Now, everything that’s made is software or has software. Everyone I talk to? Even if they make hardware? Their competitive differentiation is product management for software. In order to be a relevant company today, you’re building technology, but fundamentally what you’re doing is making products. What’s going to happen is you have to make the right products.”

2. Product management = part science, part art

Suhail said that there’s room for intuition in building products, so long as you know when to let data drive you. Great product managers know when to lean on intuition vs. data. And when it’s that latter category, PMs should welcome data as a substitute for intuition—the process of experimentation is rewarding in its own right. Suhail said:

“Product is one part science, one part art. When you think about your v1, information is limited. So, you probably have to figure out which way to go on your own. But as PMs, it’s really important to iterate, iterate, iterate. If you don’t have Steve Jobs-like intuition, you know product is a science, and it’s about iterating. The other benefit of metrics? It’s actually really cool to figure out whether you’re right. There are few fields where you do your job, guess, and then actually know whether you’re right or not.”

3. People have done this forever with less, and facts beat opinion

There was a time when people made products without data, and it didn’t really work. At Windows, Steven and his team hacked together a way to collect data at scale, even though it was incredibly laborious. Even in the pre-internet era, though, it was a powerful experience to witness data. Steven said:

“When we were building the first release of Microsoft Office (before the internet—there was a time), we built a word processor ourselves. But we were a bunch of programmers who didn’t write books, so we decided which features to put in, based on our limited experience. The way we solved our problems was with arguing—we called it testosterone-based engineering.

“Then, some brilliant people on the team decided to make a special version of Word that had instrumentation in it. There being no internet, we had to sign up volunteers, and then we would put the special version on a bunch of floppy disks and mail it. The volunteers would install, and then we would have to get on a plane and visit the computer that had version and to take all the data.

“All of the sudden we realized we were killing ourselves putting in fifty different templates, but everyone was using the default mode. OR, for one thing, we had just figured out that all Word Documents had the same typos in them. That’s where Autocorrect came in.

We instrumented. And it was a fact. It wasn’t an opinion anymore.”

4. Only innovate and you’ll lose

Everybody wants to ship a bleeding edge product that changes the world. And there’s time for that. Indeed, without it, you’d have no company. But to only do that isn’t just unwise—it’s suicidal. Suhail reflected on why the impulse to build “shiny stuff” was often a misstep. He said:

“When we work on products, we generally care about building the most innovative thing possible. After a year or two, you want to keep making innovative stuff—shiny stuff. One hard thing about product planning is to actually say no to the shiny stuff and to grind through the versions that you need.

“Guess what happened as [co-founder] Tim [Trefren] and I kept building shiny stuff? We started to lose. We started to lose customers constantly. It took us about a year to come back and double down on product and iterate, iterate, iterate. There’s a product almost all our customers use—Funnels. We’re on the ninth version.”

5. Product management’s next big opportunity? The old guard

When Suhail and Tim founded Mixpanel in 2009, even Silicon Valley was still looking at outdated, vanity metrics. But then startups began to wake up and track the meaningful actions associated with their products. Now, other industries are waking up. Steven agreed that those companies would build out product management organizations, but that they’d have carefully balance what was right for customers and the established systems. Steven said:

“They’re all worried about how their products differentiate them from existing competitors. In the banking world, for example, it turned out that having more branches was not interesting to millennials. But having an app where they can do all the same services at eleven o’clock on a Wednesday is. And having that data to drive that app well is a huge competitive advantage.

“All of these companies’ existing customers want a mobile app from them. That sounds easy, but ten usually their mobile app looks exactly like their org chart. Step One is recognition of this problem. They have to bring in someone who can look holistically at the customers and the constraints that you already have internally.”


It doesn’t matter whether you’re a leader in a lightning-fast startup or in a major titan of industry.

For both, product management is the competitive difference in today’s market. What all product managers have to learn is a discipline of data. That doesn’t mean letting data govern every decision, but knowing where it fits in your organization’s decision-making and having the patience to build something that meets your customers where they live.


Get the latest from Mixpanel
This field is required.