Google on Innovation

Marissa Mayer is Vice President of Search Products and User Experience at Google. John Hunter from the Curious Cat Management Improvement Blog dropped by Think Differently!! and left a comment alerting me to the fact that Mayer gave a talk on innovation at Google at her alma mater, Stanford University – and the talk has made it to YouTube.

As John notes, Mayer’s talk revolves around 9 key points:

1. Ideas come from everywhere
2. Share everything you can
3. You’re Brilliant. We’re Hiring
4. A license to pursue dreams (Google’s 20% time policy)
5. Innovation not instant perfection (get something out there, get feedback, and iterate quickly and often)
6. Data is apolitical (use data based decision making)
7. Creativity loves Constraints
8. Users not money
9. Don’t kill projects, morph them

Mayer’s discussion was fascinating. For example, “you’re brilliant, we’re hiring” was selected after Google put this on an internal bulletin board – and noted that this slogan got 5x the click-throughs of other alternatives posted.

The license to pursue dreams was particularly interesting. Google’s policy to allow people 20% of their time to pursue their own independent projects is well known. Mayer, however, pointed out that this time is not always or typically taken on Fridays (or other nominated day) on a regular basis – often people work on their core work for months and then spend bursts such as weeks on their independent projects. And this independent work – because it harnesses employee passion and drive – leads to 2.5x greater productivity. Mayer studied all products released in the second half of 2005, and determined that 50% of all product releases were generated as the output of the 20% independent project time. For businesses considering mirroring Google’s policy within their own organisations, that’s a powerful statistic right there.

After noting that Google prefer never to launch a product or service at the end of the week (they only do so between Monday and Wednesday), Mayer recounts a story where the team had produced a product release (Google news) – but it was ready on Thursday and therefore would not be released till next week. The team of 6 was deadlocked regarding whether to use the time to build a new feature to filter the news by date, or to filter the news by location. The team could not agree, so Mayer made the decision to hold off and release the product without either feature. After release, the comments and requests came in – around 300 requests for filtering by date, and around 3 requests for filtering by location. Goggle released an ‘imperfect’ product – but iterated rapidly based on customer feedback – a great model of user-centred innovation.