It was originally custom built in their webhosting company, but now they've created a standalone product out of it. NiceReply has API, so you can integrate it in any CRM you're already using. It's really cool, check it out.
I traveled to China this summer, and in the Shanghai airport they use a very similar system to get quick feedback on customers' satisfaction with the customs agents. At each station, there's a little box with 3-4 buttons ranging from :-) to :-| to :-(. (For example: http://bit.ly/aoSXgP)
If you're willing to structure experiments like this to limit risk but be deep enough to yield genuine benefits, there's a whole lot you can learn. You've got to work in a culture that tolerates failure, obviously, but why work somewhere that doesn't?
Back during a horrifying phase of my life, I worked in the shudder-inducing industry of internet marketing. I wanted to try a new type of lead capture form that was less user hostile. It was extraordinarily simple, but between bureaucracy and a web development group that was sorely overworked, it took 12 weeks for the idea to work its way through the constipation of the organization and into existence.
And it didn't work that well. That earned me a bit of ire.
I gave up on finding something that did work and took another, less depressing job.
The post isn't a tribute to the freedom to fail. It's about having an idea, not complicating it, and executing the basics that matter in 24 hours. There's so much you can get done in 24 hours if you don't make it take longer than that.
Of course you don't see it that way, you're just doing what you want to do, because you can.
"It's about having an idea, not complicating it, and executing the basics that matter in 24 hours" ... which is the freedom to fail.
Three things bought you the freedom to fail:
You're the President, so you have deep influence on the allocation of resources.
You value minimalism, so you limited the scope of the project to make it manageable. This decreased risk (while maximizing ROI, bravo).
You're open to an outcome that entirely invalidates the tool since you'll learn something either way.
Failure (and success) requires resources, investment and tolerance of risk. You demonstrated all three. That's the value of the freedom to fail and it's solid gold.
You don't see it that way, perhaps, because it's just how you roll at 37signals. Ask a puppy: are you trying to be really, really cute when you're yawning or are you just yawning because you need to? It will insist, as you have, that it's just doing what it needs to do.
The thing to keep in mind is that, sadly, not everyone can access the resources or commit to the investment required to get things done. Even if they do, cultural risk aversion may still paralyze them. I looked at your post and what immediately struck me was "Damn, that's good stuff. Some places are incapable of delivering this kind of thing. Why? Oh. No freedom to fail."
I guess the best luxuries are the ones you don't realize you've got. :)
It also occurs to me that committees add complication to things all the time. Being able to process an idea into a product while bypassing a committee is important factor in the investment component.
It was originally custom built in their webhosting company, but now they've created a standalone product out of it. NiceReply has API, so you can integrate it in any CRM you're already using. It's really cool, check it out.