17 August 2011

How do forecasts fail? Let me count the ways…

The only certainty about forecasts is that they will be wrong. So, let’s start to enumerate some of the reasons forecasts are wrong:
  1. Frequently, they aren’t forecasts at all; they are only guesses
  2. As W. Edwards Deming said, “Wherever there is fear, you will get wrong numbers.”
  3. Forecasts are not prophecy and were never intended to give the actual answer to, “How many units will be sold next month?”
  4. Forecasts are always based on assumptions and frequently the assumptions are wrong
  5. Forecasts attempt to predict variable behavior into the future
  6. Forecasts cannot take into consideration every potential variable that might affect the result for two reasons: a) we don’t know all the variables, and b) even if we did, there isn’t enough computing power in the universe to take them all into consideration
  7. Forecasts almost always are given as a single number (in the business context) when, in fact, they should be expressed (at a minimum) as a number and the standard deviation surrounding that number
  8. Many forecasts originate with salespeople
  9. The salesperson provided his “best guess” forecast
  10. The salesperson provided what he/she thought he/she could sell
  11. The salesperson provide a “safe number,” so that he/she can “hit her target”
  12. The salesperson provided an “average” calculated from who knows what—last year? last three months? extrapolated from last week?
  13. The salesperson provided a “big number” and will try to hit it because of pressure from sales management
  14. We try to forecast too far into the future—say, three months instead of three days
  15. The forecast is based on the assumption that—except for the things we specifically know will change—everything else will remain the same (but it never does)
  16. Our suppliers’ lead times are unreliable, so we don’t know the actual period our forecast needs to cover
  17. We forgot to carry the 2 when doing the math
  18. Our Excel™ spreadsheet formulas and references are off, but we haven’t noticed it yet
  19. Our forecasts for our finished goods are pretty good, but when MRP blows down through our multi-level BOMs, the forecast explodes due to our minimum batch sizes, percent-over and other production policies
  20. Some departments just can’t get their homework done on time and we have to produce a number from somewhere
Go ahead, feel free to leave your comments adding to the list.

[Cross-posted a Kinaxis Supply Chain Community]

No comments: