Marissa Mayer: I thought Google had a 2% chance of success

October 3, 2011, 11:43 PM UTC

By Megan Barnett

Google's Marissa Mayer, center, and Lululemon CEO Christine Day, right

FORTUNE — Several things drew Marissa Mayer to Google in 1999, when the young engineer was weighing offers from several startups during the dot-com boom. One of them was the fact that she would be the first female engineer among a team of eight, as opposed to being the first female among 48 engineers at a different startup – the jeers she received when she visited that office were enough to turn her off.

Another reason was that she believed Google had a 98% chance of failing. It turns out that 2% odds of success were far greater than what she bet on other startups at the time.

Mayer, who is now vice president of Google (GOOG), joined Lululemon (LULU) CEO Christine Day on a discussion panel that kicked off Fortune’s Most Powerful Women Summit in Laguna Niguel, Calif., on Monday. The topic was innovation, but the wide-ranging conversation touched on everything from what makes a meeting work best (the fewer people, the better) to why Lululemon stores aren’t painted beige (“We’re not a beige company,” Day says).

Mayer, 36, joined Google when it had just 20 employees, and she now helps run the technology giant, which is 29,000 people strong. She credits Google policies such as its 20% rule – employees are encouraged to spend 20% of their time developing new ideas – with its success in fostering a creative and innovative culture as its workforce has grown. Approximately half of Google product launches originate during that 20% time, Mayer says.

The key to success is often failure. “It’s totally fine to fail, you just have to fail fast,” Mayer says. Google focuses about 70% of its time on its core business, 20% of its time on related businesses, which leaves the remaining 10% of its efforts on far-flung ideas. The key is to avoid investing too much time and resources into any idea before it’s proven to be successful.

One example of a far-flung project at Google is the driverless car. It’s been met with skepticism – what does the search giant have to do with solving the most vexing transportation problems? Google has logged thousands of miles in the driverless car, but still it’s a long way from being deemed successful.

“Are we betting the farm on this? No, it’s a small team,” Mayer said. The key to success – especially with ideas in the “far-flung” category – is to make small investments with a few people. That way, innovation is fostered and failure doesn’t hurt as much.

Both Mayer and Day agree that it’s important to create an environment that empowers employees, whether you have 29 or 29,000 of them. Ideas come from the bottom up instead of top down.

One company they both admire in this respect is Apple (AAPL) – an interesting admission for Mayer, considering the fierce competition between Apple’s iPhone and Google’s Android operating system. She notes that Apple is known for what she calls “castle builder” innovation – work under a shroud of secrecy and launch a new product with a worldwide “wow” moment. Google, on the other hand, operates more like a “bird walk,” launching products early and often.

Both strategies have their own risks — and both have proven to be successful, even though they’ve each produced failures along the way.

Check out our additional coverage from Fortune’s Most Powerful Women summit.

[cnnmoney-video vid=/video/technology/2011/10/03/t_google_privacy_marissa_mayer.cnnmoney/]