Somehow I made it through a decade in tech without reading The Lean Startup. I suppose at a certain point you read about a book in enough tweets and Medium articles to feel as if you've read the whole thing. Here’s where I net out having finally gotten to the source material.

To start, there’s a lot I like about this book. Its adoption of lean manufacturing principles, its relentless focus on learning and data-driven thinking, its focus on building something people want. Those are learnings I’ve tried to keep in focus throughout my own career in product, and The Lean Startup is a great reminder to double down on those parts of product management.

But there’s also something that The Lean Startup misses, and that’s that you only get one first impression. When you’re in your early network and showing people a beta product, iterate and show them an incomplete product, sure. Find which features add value and which don’t. But when you launch your product to the world, your first complete product that’s supposed to solve a problem for real-world users, you had better get it as close to perfect as you can. It’s an idea that Stewart Butterfield talks about in his memo We Don’t Sell Saddles Here. With a first product, you only get one opportunity to show users a new way of doing things.

In Butterfield’s words:

“The reason for saying we need to do ‘an exceptional, near-perfect job of execution’ is this: When you want something really bad, you will put up with a lot of flaws. But if you do not yet know you want something, your tolerance will be much lower. That’s why it is especially important for us to build a beautiful, elegant and considerate piece of software. Every bit of grace, refinement, and thoughtfulness on our part will pull people along. Every petty irritation will stop them and give the impression that it is not worth it.”

To put too much faith into any one dogma as a product manager is to set yourself up for failure. This is often the case with acolytes of the lean startup school—they compromise on end user experience for the sake of learning. You’ll certainly learn as fast as possible, but often at the expense of a cohort of customers that will be hard to get back into your product. If your company is built on recurring revenue and the long-term value of those customers, that’s a big cost. Lean too far the other way, focus completely on end user experience but forget to measure anything at all, and you’ll end up delivering a beautiful product that no one wants to use.

So what to do? As with most things in product management, the answer sits somewhere in the middle. When you embrace lean startup principles, counter-balance them with the flawless execution Butterfield talks about above (or limit the impact of a lean release with a smaller beta cohort.) Consider data-driven thinking and experimentation, while also weighing the intangibles of what makes a product great. You have to see all of these things at the same time to make the right decision.

One of the first things you learn when you’re starting to drive is to keep your eyes moving. You can’t just stare straight down the road, the tunnel vision will keep you from seeing what’s coming into view. Instead, you keep your eyes moving, a dart to the left, to the right, to the rear-view window, to the speedometer, back to the road. It keeps you alert. You see mistakes before they happen. The Lean Startup is one part of the road to keep an eye on. Your dashboard. A valuable tool, just remember to lift your eyes up from the numbers every so often to see where you’re going.