You aren't measuring things that really matter.
There can be lots of reasons for this issue. Perhaps you are measuring something that has little or no impact on your competitive position. In strategic planning, this is worse than a waste of time.
Ask yourself: if ONLY this number improved, would we really be better off? In many cases, the numbers you look at overlap with other numbers. To use a simple example from financial measurements, it's somewhat redundant to look at both gross margin and net profit, because net profit is simply gross margin minus fixed costs and a few other expenses. I won't tell you which you should look at in your own company, but I will tell you it's rare that a company should include both numbers in the five to ten core metrics in a streamlined balanced scorecard. This means you need to pick one, understand that it (like all metrics) will have the flaw of not being complete, and move on.
Another reason why companies sometimes measure things that don't really matter is that the metric chosen makes the company look good. At worst, this can be a case of throwing in a number because it can be counted on to give us good news even when the other numbers are telling us bad news. Again, if the number does not affect your competitive position or your ability to succeed, consider eliminating it from your scorecard.
There is an even more insidious problem associated with measuring things that don't matter - if this is your company's problem, you are very likely shying away from measuring things that matter very much. Sometimes this is because managers fear the implications of managing to certain metrics. If managers fear measuring profit per employee, for example, because it might lead to staff reductions, you should be asking yourself whether this fear is useful. There are many industries where staff reductions are a vital strategic management tool - failing to consider them when appropriate can be a major strategic error.