Groundhog Day Results – Forecasting Accuracy Needs Innovation

Groundhog Day 2013: The results are in. Punxsutawney Phil, the pesky rodent with 15 minutes of fame every February 2 since 1886, did not see his shadow this year, and so we are looking for spring right around the corner…  Maybe.  Groundhog Day festivities have their roots in European weather lore, where a badger played the part of the modern day

Groundhog Day Forecasting


groundhog.  My personal tie to Groundhog Day is much closer to the 1993 comedy starring Bill Murray.  But behind the fun and festivities lie some somber statistics.  According to the StormFax Weather Almanac and records kept since 1887, Punxsutawney Phil’s weather predictions have been correct only 39% of the time.

Industry hopes to do much better with forecast accuracy than 39%.  Statistical models are invaluable in building confidence and success for companies struggling to compete in the global marketplace and vie for the dollars of a customer in a recession squeeze.  So many factors are out-of-control for the executive today, who deals not only with weather disruptions from the statistical norm, but many other outside factors:  loss of revenue channels, bad relationships and impacts from other links in the supply chain, and a lack of supply chain talent in their own organizations.  Executives then look inward to best practice technologies to attain the best possible supply chain results possible and thereby to weather the storm.

As supply chain technology evolves, the better tools move beyond statistics, and with neural networking can learn as they go, drawing from sources like their customers’  POS data, current market share, and unemployment figures.  They take the team to the next level of maturity, still accepting the input of the professionals, but not leaving any available external information on the table.

Superstition, if you want to call it that, can block progress when management does not want to relinquish control of the data to the powerful new methodology, and defeats the purpose of investment and training in the tools that were there to help in the first place.  The best S&OP software will have the transparency and ease-of-use to allow the whole organization to overcome those fears.  It will be backed with dedicated education and training to make it quickly become a part of the company culture.

Superstition can also divert selection of best-of-breed software to large enterprise software, with a feeling of safety that the larger and more all-encompassing the software, the better.  It may feel like a safer, easier, and less controversial choice.  It may have a weaker forecasting engine than other best-of-breed software.  It is less work in the short run of the procurement process, but also offers far less gains in the long run of the lifetime of the supply chain.

Without powerful forecasting, businesses can still rely on operational improvements to guard against supply chain losses:  adding supplier diversity, forming alliances with like businesses to fill holes when inventory predictions come up short, using local supply chains for local demand, and consolidating distribution centers.  But imagine the power of combining all of these far more costly operational improvements with the software that provides the best forecasting accuracy available on the market.


Another relative of Punxsutawney Phil’s is Gnocchi the Squirrel, who by virtue of eating more peanuts out of a Romney bowl than an Obama bowl, predicted a win for Romney last November.  More superstition.  Demand Foresight uses innovative technology that is unlike any on the market, and partners with your company to take your forecasting accuracy out of the shadows!

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title="" rel=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>