Incentive plans are designed to reward specific behaviors, but they're often ineffective at best and harmful at worst.

Here are five common ways employee incentive plans can produce the wrong results:

1. You foster the wrong outcome.

What you reward drives very specific behaviors, often with unintended consequences. Say your goal is to increase sales so you reward salespeople for generating revenue.

Fine--but what kind of revenue? If your salespeople have reasonable price discretion, it can be very tempting to cut the price to the bone, land the job, book the revenue, book the commission, and let someone else worry about how profitable the job is.

If you're struggling and you're desperate to keep the lights on, maybe you'll take any revenue you can get. Otherwise, what you really want is profitable revenue. Reward salespeople on profitability, not just gross revenue. Then they have an incentive to sell a lot of work and to sell at a price that helps you turn a profit.

2. You cap performance.

Incentive caps eventually de-incentivize outstanding performers--or at least limit their overall performance.

Say your salespeople are on a 1 percent commission rate for sales up to $500,000 per month. Sales over $500,000 don't earn a commission. Next month they start over.

Why did you cap commissions? Maybe because you don't have the capacity to handle additional sales (although that's a poor excuse when you think about it.)

So what happens? An outstanding salesperson makes sure she goes ever so slightly over $500,000 each month. She could sell more--but what's the point? If she hits her cap on the 25th, she might decide to hold off trying to close the deal with a hot prospect until the first of the month. Or she might just relax a little and prepare to hit the ground running on the first.

Incentives signal desires--so even the best employees soon decide that since you don't want to pay for sales over $500,000 a month, you clearly do not want sales over $500,000 a month. So they won't give them to you.

And, if somehow they accidentally (because it will be an accident) do sell well over $500k one month, they definitely resent the fact there is no additional reward.

The reasoning behind most commission caps is based on some form of, "I don't want to pay anyone more than that." If the results are what you want, pay for them. $200,000 in additional sales is worth paying a $2,000 commission.

If it's not, you shouldn't be paying commissions at all.

Quick caveat: Say sales in excess of $500k create operational burdens like overtime or other increased costs. If that is truly the case, consider sharing the "pain" and reward incremental sales at .5 percent. Or allow salespeople to roll some percentage of additional sales into the next month.

And then work your butt off to improve operations so you can efficiently handle additional volume--because where sales are concerned, too much is never enough.

3. You create an imbalance between reward and risk.

When there is little downside and too much upside, the risk-reward equation can get seriously skewed.

Say the bulk of your revenue comes from two or three key customers and you really want to land more customers. So you add an incentive to your existing 1 percent plan that pays out a 10 percent commission on sales to new customers.

At face value that's fine, but if the bulk of your salespeople are currently earning a competitive wage they have little incentive to grow their current customers and every incentive to try to land new ones. The downside risk (maybe losing out on some 1 percent sales) is nothing compared to the upside (10 percent on new sales). That means they'll call on a lot more marginal prospects, maybe worry less about building a relationship than about closing a deal at almost any cost, maybe create friction with your support employees as they try to pull strings to swing deals, maybe start to pay less attention to their current customers--and suddenly the culture of your sales organization, and your company, changes.

Good incentive plans balance rewards with risks--both for the employee and the company.

4. You create conflict.

Conflict is hard to avoid, especially when interests between different functional groups aren't aligned.

Say the goal of your accounting team is to invoice within two days of job completion. All they care about is closing the job so they can invoice it and hit their incentives. Your ops people care about closing jobs but not really: They care a lot less about yesterday's jobs than they do about running today's work as productively as possible, since that is the basis of their incentive plan. So if their paperwork is incomplete, "We'll work on that soon." Or if they forgot to code certain activities, "Hey, sorry about that, we'll fix it tomorrow--right now we're running product."

You can never totally eliminate conflict, but you can work to optimize operations so that conflicts are less likely to occur. Decide what you want--then work hard to make it as easy as possible for employees to deliver what you want, whether on their own or cross-functionally.

5. You establish too little line of sight.

The key is to reward specific behaviors employees control--not the results of factors employees can't control.

In the 80s (or 90s, can't remember, I'm old) at R.R. Donnelley every employee was given stock options: If the stock price reached a certain level by a certain date, we all won. The idea was to help us think big-picture, to pull together, to feel we shared in company success, etc.

The problem was that stock prices, especially in the short term, typically rise and fall due to factors far removed from the average employee's control. In our case, commodity and supply prices could rise, overall market demand could fall, e-books could become a glimmer in some smart developer's eye and the market could perceive that print is dying. So, a guy working in a manufacturing plant was a lot more like a cork bobbing in the sea than a captain of his stock price ship.

So what can happen? You work extra hard, yet the stock price still falls and you decide your individual effort doesn't really make a difference.

In most cases the better approach is to reward specific behavior, not general results. If you feel increased productivity is a worthwhile goal, reward production employees for faster job changeovers or increased up time--reward them for hitting productivity targets. Or for performing specific actions that generate cost savings.

Pick a behavior that creates a desired result, and reward that behavior. That way I can tie my reward to my actions.