'

The power of hobbits giving 20 percent

Every new essay by Paul Graham on startups is like a chapter of a Tolkien book, telling the long and winding story of how the powerless can change the face of the world through the simple action of believing in their own abilities.

Every new essay by Paul Graham on startups is like a chapter of a Tolkien book, telling the long and winding story of how the powerless can change the face of the world through the simple action of believing in their own abilities. Graham's latest, entitled The Power of the Marginal, explores the advantages of being disenfranchised.

Being "eminent", as Graham calls the state of having been successful in the past, is in his world like a deadweight around the shoulders of a once-vibrant innovator. As I said in my last entry, this message is one I fully support. But if you're a manager or programmer inside a large company, by Graham's definition you're an insider. You suffer from the same timewasting responsibilities and vested interests in soon-to-be-outdated practices and products. How do you break free of the shackles of eminence?

Google's famous answer is 20 percent time, where their engineers are instructed to spend a fifth of their time working on their own projects without management oversight, which might then be subsumed into the Google machine. Would this work at other big companies, particularly non-tech businesses? Maybe. Would your bosses have the cojones to actually implement such a policy and give up 100 percent oversight control? I'm guessing not. The way this would most likely work is on a case-by-case basis. You would go to a manager and give him or her a proposal to use X percent of your time discovering how a particular buzzword-compliant sliver of Web 2.0 could be implemented in an hour here, an hour there while the main part of your working day is filled with listening to overpaid consultants lecture you about how to implement ERP, CRM, BI, or whatever other white elephant is on today's menu.

You'd need a compassionate manager to agree to such a dangerous plan, and they would need to have faith in you to spend that time wisely, even if not successfully. And, as the above Joe Beda link shows, it works best when all engineers are given the same freedom and, crucially, they can brainstorm amongst themselves where there is no hint of power differentials between those sharing information.

The Google culture is one you would not see replicated at any non-tech company of their size. That doesn't mean you can't at least try some of their ideas in miniature. As Tolkien told and Graham echoes, power can be held in the smallest of things.