Bonuses Better Than Rasises

Bonuses generate improved productivity more than raises do, according to a recent study from Cornell Center for Hospitality Research, reported in Science Daily (hat tip to Guy Kawasaki.):

"Giving a 1 percent raise boosts employee job performance by roughly 2 percent, but offering that same money in the form of a bonus that is strongly linked to a job well done can improve job performance by almost 20 percent, finds a new Cornell study on the relationship between pay and performance."

OK, you’re smart enough not to through out your entire compensation system because I quoted a summary of the research.  But stop and give some thought to your overall incentive system.  Unless your business is really unusual, your incentive system copies what most others do.  If it has incentive components, they aren’t well thought out, and they are gamed by sales people and managers.  So stop and take a deep breath, asking yourself if you are really incenting the behavior you want.  Here’s a short article on incentives I wrote a few years ago that might prove useful:

On a recent trip to London, I heard the story of Jeremy Bentham and Edwin Chadwick and the Australian prisoners, which offers a lesson to all managers responsible for incentive systems.

For many decades in the 19th century, England sent its convicts to Australia. Edwin Chadwick was the first public health officer in Britain, and he was concerned about the low survival rate on the ships carrying prisoners. One day, while walking along the Thames, he ran into his mentor, economist Jeremy Bentham. As the two men watched the loading of a prison ship, Chadwick remarked that only 50% of the men would survive the trip to Australia.

Bentham asked about the contract that the government had with the privately-owned ships, and Chadwick explained that the ship owner received payment for every prisoner who boarded the ship. Bentham suggested a minor change: make the payment per prisoner who disembarks in Australia. On the first voyage under the new terms, the survival rate was not 50 percent, but 98 percent!*

The first question that a manager has to ask is what behavior he or she wants from the people being incented, be they employees or contractors. That is, do you want prisoners put on board in England, or prisoners unloaded in Australia? That sort of simple question is mangled badly time and time again. Many of the telecommunications equipment companies last year incented their personnel to show strong quarterly sales growth. At many companies, they got the growth – but with substantial discounts and extravagant credit terms. The result was sales without profitability. The opposite tack was taken by banks in major mergers a few years back. They wanted high profit margins, and incented their managers to maintain margins. The executives got what they want, for a little while. Margins held, but customers fled. In both cases, the problem was that incentives did not favor what the company really wanted: a stream of profits that continued over time.

Solutions to bad incentive systems are not always as simple as Bentham’s. They may require updated data systems, approximations, and tradeoffs between conflicting goals. But even if the perfect system is not easily obtained, improvements on bad systems come quite readily, and provide great value.

*Thanks to John Blundell of the Institute for Economic Affairs in London for the story.