Long ago, W. Edwards Deming warned us about the use of what he called “arbitrary numerical goals.” Targets are another name for these. They are so commonplace that governments, service, and manufacturing organizations all use them. Targets have become accepted in all organizations, but this habitual use conceals the harm they actually do.
If you are an executive, middle manager, or worker, odds are you have targets to hit. Many times these have to do with sales, budget (financial), output, or anything else that can be dreamed up. Targets seem to make sense because they give all involved clarity over things that they can “control.” Additionally, targets provide feedback and achievement: If I hit or exceed the number, I have done well (maybe even get a reward), and if I don’t hit the number, I get paid attention to by someone above me. This is true no matter what level you have achieved in an organization.
In the book, Freedom from Command and Control (Productivity Press, 2005), John Seddon outlines many of the arguments one might hear about using targets:
• Targets motivate people.
• Targets set direction.
• Targets should be based on what is reasonable.
• People should be involved in setting their targets.
When sales are involved, rewards often are given in achieving targets. Conventional thinking tells us that this motivates people. Indeed, it does: It motivates people to manipulate (even cheat) to hit the target(s). This may come in the form of overselling a product or service that is later returned or cancelled. I’ve seen this far too often in functionally separated organizations when the sales department has sales targets, and the operations department gets to manage costs. The result is called “American toast.” In Deming’s words, “You burn, I’ll scrape.”
As Seddon said, when targets are set in any part of the organization, they become the de facto purpose of the organization; meaning workers, managers, and executives are focused on meeting the target and not the customer. A lack of focus on the customer and forgetting why the customer chose to deal with your organization leads to waste, but when you ignore the customer in favor of targets, the loss is massive.
Some argue that targets set a direction, but isn’t it enough to know that we want sales to go up and costs down? What we need to concern ourselves with is the method, not a target. If we are making American toast, we should look at end-to-end costs of hitting revenue targets if costs outstrip the gains. Revenue and costs are two sides of the same coin.
Some chose targets based on such nonsense as SMART, which stands for specific, measureable, attainable, realistic, and timely. This approach for setting objectives seems reasonable but tells us little about what the right measures are for an organization. However, they are still arbitrary. The correct measures must be gained from understanding systems, and they will require no targets. This is true even if you set your own targets.
Those of you who understand variation can appreciate where Deming is coming from—or you should. Look at the control chart below. It represents sales by week (in thousands). The lower control limit is 47.4, the upper limit is 182.9, and the average is 115.1. Because the data represent common cause variation (between the limits), the system that produces these sales will continue to operate between 47.4 and 182.9 unless the system (or method) changes.
Click here for larger image.
I’ve seen too many managers look at data without understanding a basic control chart. I’ve seen some executives, when shown a chart, try to set the target at the average (in this case 115.1), which of course will lead to you hitting the target about half the time and missing about half the time. Others try to set the target at the lower control limit, which means that you will hit your sales target all the time, but won’t improve the system. Still others will “raise the bar” or set “stretch targets,” which still doesn’t change the system.
No matter where I set the target, the system will not change until the system that produces it changes either positively or negatively. If organizations understood variation, they would never set targets; they would (hopefully) use the data as a performance baseline and improve by experimenting with method.
In my next column, I’ll talk about an approach to establishing measures rather than using targets. There’s a better way if you are prepared to change the way you think about the design and management of work.