Tom Reilly Quoted in new book

Shaun Snapp has a new book out titled "Supply Chain Forecasting Software" available here.

In the Chapter titled "Why Companies Select the Wrong Forecasting Software" on Page 211 of the book, we are quoted as saying:

"Companies choose Integrated systems that provide full ERP solutions based on the old “you won’t get fired for buying IBM” herd mentality. These large ERP womb to tomb solutions are the market leaders, but do each of their components lead the market? We think not. IT drives a lot of the decision making and they have a perfunctory view.” The “one size all” garments never really do fit quite perfectly. The road less traveled can be rocky, but can also be rewarding.

The forecasting systems found in the major ERP systems use simple and flawed models that don’t question the data and end up with silly baseline forecasts. They ignore the concept of level shifts, seasonal pulses and changes in trend. The way they handle outliers is far too simple. If you give them a time series like 1,9,1,9,1,9,1,5 they won’t be able to find the outlier or in this case inlier.

You need to consider why the large ERP vendors don’t participate in forecasting competitions. Why should they risk their market position? Why should they be exposed when they are fat and happy? You should consider reviewing results from independent competitions to see who is delivering quality forecasting tools Many who participate are not software houses, but they are a few that did."

Shaun is VERY knowledgable about forecasting.  He addresses many things we find important to us.  With that said, Shaun and Autobox have not had a good introduction yet as Autobox and may of the ideas around "withholding" data which we are flat out against and light discussions on outliers that we feel could be enhanced.  I have sent my comments to Shaun that made him want to engage us more and he offered for us to expand on my points in revised versions of the book.

Mike Gilliland from SAS gave it rave reviews.



Go to top