Danger: What you know about compensation may be only half-true

Is your compensation system working? That’s a tough one. After all, reliable evidence of comp-plan effectiveness is hard to come by, say Jeff Pfeffer and Bob Sutton, Stanford professors who have studied and written about compensation and rewards.

Without hard data, senior executives are often left relying on preconceptions, untested assumptions and experience-based biases about what works or what “must” or “should” work.

When results aren’t there
Sometimes, these notions make sense. But often they amount to half-truths that can leave execs wondering why a comp plan isn’t getting the results they want.

Possible consequences:

• Turnover. Star performers leave, but mediocre performers stick around.
• People do the minimum. The next generation of potential stars get discouraged and “retire in place.”
• Unexpected side effects. People’s efforts don’t seem aligned with company goals.

Half truths
So what are these half-truths and assumptions? In their book, “Hard Facts, Dangerous Half-Truths & Total Nonsense: Profiting from Evidence-Based Management,”Pfeffer and Sutton point to the following:

1. People are motivated by money. To a degree, employees are (especially lack of money). But different people are motivated by different things. Some are driven by opportunities to achieve tough goals, innovate and build a business. Others are motivated by teamwork and friendships with co-workers. Still others just want to collect a paycheck and get home.

Key: Identify an employee’s biggest motivation. If it’s money, then you know how to motivate them. But what if comp is second or third and what they crave is individual attention from an experienced mentor? Give them that – if possible – and make sure the comp plan doesn’t demotivate them. (That is, companies can and have overplayed the non-monetary compensation card.)

2. Individual performance can be reliably and accurately measured. Sometimes you can measure individual performance. But it’s usually not easy. Some issues: Do performance reviews include subjective judgments? Supervisors may use differing scales for things such as “cooperativeness” and “attitude.” Or if you decide to get completely scientific in your measurements, then what about human “intangibles” necessary for success?

Key: When you’re basing raises on performance reviews, recognize the difference between reliable measurements and subjective (thus less reliable) ones. Make sure that those measurements also accurately reflect what’s necessary for job success.

3. Performance is a solo activity. Managers want individuals to be accountable for their success or failure. At the same time, people may be dependent on others for success. They may depend on the information other people give them, the ability of other people to articulate problems to them, the industry, the larger economy, corporate priorities and organization competence, and just plain dumb luck.

Key: Make sure your comp system doesn’t “pretend” people are independent of outside issues.

4. What worked at my last company will work at this one. Maybe. But given the difficulties mentioned above, will a new comp system based on a different company really work? Is the industry the same? How much time has passed? Is the market different? Is it clear that previous comp system delivered the results cited, or were there other factors involved?

Key: Before revamping a comp system, conduct a small test. That may allow you to adjust the system to reduce unwanted side effects.

5. Consultants can figure it out for us. Some consultants deliver. But Pfeffer and Sutton cite a distressing example: A senior HR consultant admitted to them that his company pushes pay-for-performance plans, even though these almost always fail. That’s because consultants get paid for advice, and maybe to implement a plan, but not for business results. Worse: When the plan flops, the consultants get paid for trying to fix it.

Note: No system will work without managing expectations. Based on regular supervisory feedback, people should know about how much of a raise to expect come raise time.