Innovation and the 20% Solution

As IT budgets threaten to follow the same trend lines as financial markets, it's a natural impulse for managers to circle the wagons, concentrate on core projects and put off innovation for another day.

Natural, but wrong.

For proof, look at what happened after the Internet bubble burst earlier this decade. Sure, that tech bust pales in comparison with the current worldwide credit crunch in terms of overall effect, but if you consider the Internet sector alone, the money drought and corporate failures were pretty stunning.

So, how did Google emerge from that wreckage to become not another, but a multibillion-dollar company and the world's most influential Web brand? A key part of the equation has been constant innovation (either in-house or by acquisition). It's not easy to innovate when money is drying up all around you. But Google managed to do just that during those lean years.

One important policy has increased both employee satisfaction and innovation at Google: the "20% rule," which allows engineers to spend one-fifth of their time on corporate projects of their choosing -- creating something new or making something work better -- even if the project isn't part of their job descriptions.

For one day each week, Google's engineering staffers get to work on projects they think are important for the business, not what management has prioritized for them.

Before you scoff at that as a tech-bubble luxury that only an overstaffed company can afford, look at some numbers. Google's AdSense for Content was developed as an engineer's "20% time" project. Last quarter, Google generated more than $1.6 billion in revenue from AdSense partner sites; that's almost one-third of the company's total revenue.

A few other companies have similar policies that have yielded noteworthy results. 3M's "bootlegging" rule allows research engineers to spend up to 15% of their time on projects of their choice. One well-known outgrowth: Post-it notes.

What the 20% rule has done at Google is turn a significant chunk of the company into something akin to a venture-capital innovation laboratory, but without outside funding to seed the work.

"There is a big difference between pet projects being permitted and being encouraged," Google software developer Joe Beda wrote several years ago on his blog. "At Google, it is actively encouraged for engineers to do a 20% project."

Beda outlined aspects of the environment that makes 20%-time success more likely, such as a single code base that makes it "really easy to look at and contribute to code in other projects without having to talk to anyone, get special permissions or fill out forms in triplicate," and a culture of transparency so teams share "the most intimate details of their project."

Only exceptional managers are going to buy into the idea that their most valuable assets -- their people -- will be available for company-directed work just 80% of the time. And there have been some rumblings on the Web that Google's 20% rule has come under pressure, especially if main projects are falling behind schedule.

It will be interesting to see whether Google's 20% rule can survive the current downturn. The company said earlier this month that it's shutting down a number of fledgling, experimental Web products and services.

Some of those decisions made a lot of sense, such as ending user uploads at Google Video (hardly needed, now that Google owns YouTube). Some other promising services that were shuttered -- such as a mashup editor that had been in "limited private beta" -- may have been victims of the down economy.

But great things can happen when tech workers in the trenches can spend time pursuing their own ideas. Even when budgets are tight.

Sharon Machlis is the managing editor of You can reach her at

Copyright © 2009 IDG Communications, Inc.

Shop Tech Products at Amazon