Hi there,
I am assuming that the information you have provided is consumption for the same material at consecutive periods of time (months, etc). My suggestion would be to check the following points:
- How many intervals is SAP using to calculate the weighted average (you have 9 months but SAP might be using the last 12 - I think that's the default value)
- How big is the difference between your calculation and SAP results. There might be a problem of floating point give the big variation in the consumption values (from 1 to 400-500 and then back to 30 or so). For a forecast model this variation is very dififcult to handle properly even for sophisticated systems. The handling of the error would cause significant variations in the end results. For instance, results for a 90% confidence level may vary significantly for a 95% confidence level
- As a corollary of the point above, may I suggest that you revised the consumption figures to better "explain" the big jump from 30 to 400 and back to 30. Was that an anomally on youor sales cycle? if so, you should remove the anomaly (in fact, reduce the consumption) and try again. If not, you may have to include the more periods in your analysis not only 9 to account for previous "anomalies"
This might not be the straight answer you were looking for but unfortunatelly it might require a bit of of work.
Please let me know if I can help you further or if I completely misunderstood the question.
Cheers,
Rodrigo
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.