I keep my watch about 5 minutes fast. I do it mostly to remind me that I'm committed to being on time. And, I mentally adjust for the 5 minutes, so I have a rough sense of what time it really is. Here's what struck me this morning: my watch is never right. But, it's sort of right, and good enough for my purposes. By way of contrast, I could have a watch that was - at times - exactly right: if the hands never moved. Of course, that would be useless for anything.
I think a lot of strategy processes are aiming for "really right", but end up with a false sense of confidence. The sort of analysis that underpins much strategic work is sometimes right, but it's hard to know from the analysis when it's not. Too often, the result is decisions that don't reflect the degree to which is the underlying analysis might be wrong. I'd rather have a watch I know is roughly right, than one I think is exactly right, but only now and then.