People love to talk about how "perfection is the enemy of the good," and for entrepreneurs that couldn't be more on the money. Especially in the online space, getting something perfect is a far-fetched goal that takes way too long given the quick business cycles. Instead, I've found my best successes have come from pinpointing the basic customer need (the critical 20 percent of a project's features), quickly getting a foundation product out into the market, and then iterating on it until it really hits the mark.

That means at CarGurus we usually ship products before they are "perfect." It's not that we have low standards, but rather that by resisting the urge to strive for perfection we are able to work and respond faster, ship products quicker, gather consumer input earlier in our development cycles, and ultimately arrive at a better result.

Long development cycles with an emphasis on perfection bring a lot of risk. Long ago, I was a product manager at a major enterprise software company, and we had 18 month development cycles to ship new products. We wrote detailed technical specifications, talked to customers, collaborated with developers, and endlessly tweaked things before we were comfortable with shipping the product.

In today's environment, and particularly at a consumer web company, that kind of lag gives competitors all the time they need to blow you out of the water.

The beauty of an internet business is that we don't have to make it perfect. We do have to ship products that are not full of errors and "bugs," but the initial product does not have to be perfect. At CarGurus we release new features every night to our users, and if it's good enough to engage visitors then we improve upon it. For example, last year we launched an email newsletter to update people on the latest car deals that matched their search criteria. We put the initial product out there with the minimum amount of functionality we needed. If nobody wanted that initial cut at the product, the effort to perfect it would be a complete waste of resources. We got the product out quickly with the minimum features and found that consumers loved it. Subsequently we kept iterating the product to add additional features but it was that initial fast launch that gave us the confidence and data to know that the product was on the mark.

I've learned that speed to market is the critical element in product development. I've worked with engineers who spent weeks trying to get products just right, while competitors gained ground. So I admire people who want to be perfect--I just don't hire them...