The Maximum, Beautiful Product

Editor's note: Glenn Kelman is CEO of Redfin, a technology-powered real estate broker, backed by Greylock Partners, Madrona Venture Group and CrunchFund, with more than $7 billion of home sales. He previously co-founded Plumtree Software, which had a 2002 IPO. He writes a quarterly column on Silicon Valley real estate for TechCrunch.

For years, all of us in Silicon Valley and its outposts have embraced the idea of the lean startup, shipping the minimum viable product, failing fast and iterating quickly based on customer feedback.

This has been the perfect development paradigm for a hit-driven industry, with bright flashes in the pan, but also plenty of silt. Why dedicate years of your life to making the perfect hula hoop if it might never catch on?

The problem with this paradigm is that it isn't always the best way to build disruptive technologies, which, as Michael Arrington noted yesterday, lately seem in short supply. How many big ideas "failed fast" and were discarded just because they were half-finished?

If Steve Jobs had shipped the minimum viable iPhone, might we have concluded that people preferred a keyboard? If Tesla had manufactured the minimum viable car, would anyone still be driving it?

Maybe the reason we don't have big ideas is because our entire approach to building them is sometimes so frugal with time, money, and belief. Lean startup techniques have revolutionized how we build software, but the lean startup has also turned the a startup's only initial asset – the conviction that an idea will work — into a villain rather than the hero. This is why we so often see startups pivot rather than persevere.

At Redfin, we have divided projects into sprints and fits and starts, which has led to incremental improvements, often delivered at a surprising speed. But we also try to set aside time for projects that take longer, such as our recently redesigned page for showing all the details about a listing. Of course you can still build something big, one piece at a time. But without conviction, a lot of little things never quite add up to a big thing.

In my own experience of ADD businesses filled with uncertainties and competing priorities, projects get sidetracked from one sprint to the next by a shiny new idea, or derailed entirely by user feedback on a product you already knew was half-finished.

You often never find out how good a product might have been. As I've gotten older and the world has gotten faster, the hardest thing to do has been to take the time to find that out, to create space for myself to screw around in speculative ways, to stay safe from criticism until I've hit upon something that really pleases me.

Because the truth is at least one critic has spoken long before you release your product to the world: the beast at the back of your mind always yelling "FASTER, FASTER," so loud that the little poet in there hardly has a moment to think. Even worse, sometimes the beast says "What's the point? Of working on a little detail no one will notice? Of exploring an off-the-wall idea likely to end up in the wastepaper basket?"

This isn't a repudiation of the lean startup, just a reminder that it can't work without conviction, and it isn't always the best way to do something big. Some products are a ditty, and some are a symphony, and sometimes you just have to let the music play.

[image via Flickr/Dream in the Dark of Day]