You Get What You Don't Pay For: The key to knowledge management

My friend Dan Ariely, a Professor at MIT's Media Lab and Sloan School , and his colleague James Heyman wrote a wonderful little paper called Effort for Payment: A tale of two markets.  In this short piece they brilliantly show people operate with at least two core motivations -- social and economic -- and hence live in two different "markets" for all our actions.  Social markets,  when we are asked to do favors, or work for charity, or provide for love ones, are markets in which effort is independent of payment.  If you give me a little gift or a large gift, my effort does not change.  However, once you introduce a financial incentive in a social market, my effort (in their study) actually decreases, because I am now working for money, and my effort is linked to how much I get paid.  If one continues to increase payment, one can get effort back to where it was when I was working for nothing!  Put another way, a small payment is not only a bad incentive, it pollutes what otherwise was effort that did not have a price by turning a social market into a financial one.  So what?  Well, I would argue that knowledge is usually a social good, and when an executive decides to begin to pay people for contributing their knowledge for the good of the company, it probably decreases their incentive to participate.  In my earlier blogs, Where's Your Wiki, and wikiCalc: The next killer app?, I posit that there are "altruists" who will give knowledge to the community as part of their natural interaction.  If you begin to pay these altruists, my guess is that they will be less likely to continue participating.

Interestingly, there are many different financial markets for the facilitation of knowledge creation: the patent system is central, as is digital rights management.  Most commercial organizations do not allow their employees to harvest the benefit of patent creation.  Firms claim, rightfully, that all employees work is "work for hire," and therefore the property of the firm.  There are incentives, usually social, and sometimes financial. 

I would argue, that most companies are much better off giving their scientists and any other knowledge worker, non-financial benefits for creating and sharing new knowledge.  Awards, recognition, and more intellectual freedom are the coin of the knowledge realm, not money.  If you begin to pay, the process will dry up.  In Dan's experiments he found that if he gave people a small candy bar, or a lovely box of chocolates -- they worked at the same level.  However, as soon as he said, "Here is a 25 cent Snickers bar, and here is a $5 box of Godiva chocolates" he found that work dropped tremendously for the 25 cent Snickers bar for once they knew the "price" of their effort they became underpaid. 

My guess is that if you want to keep things in a social market, and thereby keep effort independent of reward, you want to make sure that the recognition you provide is idiosyncratic and random -- so that the recipient, and those who are watching the recipient receive the prize, cannot translate the "gift" into a money equivalent.  As soon as you say the Snickers bar is worth twenty five cents, the magic of the social market is broken, and you have to break out your wallet and begin to really pay for effort. 

It is possible to create a pay for effort system -- and patents attempt to do just that -- one only needs to create an incentive high enough for the exertion of great effort.  What I think is fascinating is that most organizations can benefit from lots of local sharing which can occur without financial incentives, but rather social ones.  Moreover, firms need to realize that if they pay for performance when it comes to knowledge, they better be ready to go all the way and give significant payment for significant ideas or inventions.  Giving small payment will likely cause motives to diverge, not align.  As my grandmother used to say, "In for a penny -- in for a pound." 

Marketing Remix (with Antony Paoni)

wikiCalc: The next killer app?