Sales Forecasting

The forecasting team at LEGO uses moving forecasting.  The forecasting department uses a system called Statistical Modeling which accounts for historical seasonality as well as similar item performance.  Forecasting for LEGO is very dependent on history.  We can imagine that a City themed item with TV commercials in Easter and Christmas in 2008 will have a similar seasonality to a City themed item with TV support during Easter and Christmas for 2014.  Now the number of units sold will increase as the business continues to grow, however predicting quarterly seasonal demand will not change drastically; if at all.   In addition to our sales forecast being used to send a demand signal to our manufacturing team, it is used to set budgets.  Annually we rely on our October sales plan for the following year to set the budget for that year.  For example, in Oct 2013 we have a monthly consumer and shipment forecast for 2014 that is used to set the 2014 budget.  An example of a budget relying on the shipment plan for 2014 is our in-store budget.  I need to know the shipment projection because a percentage of shipments for my account is allocated to my in-store budget.  This will allow me to plan model boxes or additional in-store visuals that could be placed on the store shelves.  On a monthly basis we revisit actual shipments and consumer from the previous month to see if we hit plan or not.  Based on this information, we re-look at the balance of year shipment and consumer plan to determine what actions to take.  If we take the plan down, budgets such as in-store, promotional support, or the field merchandising budgets could get reduced.

Answering a question from this week’s lecture, “In Chapter 6 of Lehman and Winer, it’s stated that the most frequent used quantitative method is the moving average despite the availability of more sophisticated methods.  Why might this be true?”  Moving average forecasting is used more often because it is easily understood, easy to compute, and provides a stable forecast based on historical data.  One of the advantages of using a moving average forecast is that it smoothes out peaks and valleys in the set of data.  Moving average forecasting provides stability for an organization.

Looking at 2 of my classmate’s blogs I would like to call out Jacob’s week 2 blog and Jennie’s week 1 blog.  Week 2 Jacob did a great job relating Drucker’s idea of the difference between the needs of the seller and the marketer to real world examples.  I appreciated the JCPenny and Macy’s ad difference.  JCPenny took more a sales oriented approach as they focused on selling the consumer this one particular shirt.  On the other hand, Macy’s was more focused on selling the Macy’s brand to the consumer and ultimately letting them decide what to buy but ensuring that it was at Macy’s.  JCPenny offers consumers a button-up shirt, but Macy’s offers their consumers a full 360 degree experience offering many different products endorsed by celebrities.  Macy’s took a marketers approach on selling the Macy’s brand value to consumers.  Week 1 Jennie discussed something that companies deal with every day, the ethical dilemma around maximizing profits at the expense of safety.  In addition to this topic, I also thought it was interesting to read her experience the first week reading Drucker since her background is pharmacy and not business.  In the week 3 class videos we reviewed political and legal environments of marketing.  In this video there were federal, state laws, and regulatory agencies that protect consumers from organizational decision.  In Jennie’s blog, she mentioned that New England Compounding Center bypassed legislative decisions in order to maximize their profits.  Many safety tests were overlooked as the only concern was to sell the most units to consumers.  The end result was NECC produced a lackluster product that got many people sick and they lost all of their consumers.  Ultimately legislation protects consumers from companies, but as in NECC’s case legislation would have protected the company from itself.

We were asked this week to determine what the percentage of our research cost was to our total marketing budget within PharmaSim.  I ran 6 years worth of simulations and found that the research costs rose as a percentage of my marketing budget year over year.  In Period 0, 1, and 2 the research costs were 1.2% of my marketing budget ($440K of $34.2M, $460K of $37.9M, and $481K of $38.9M respectively).  This total dollar value of research continued to rise regardless if my sales or profits were positive or negative for that year.  As we moved into Period 3 research rose to 1.4% of my marketing budget.  Period 4 was 1.6% of the marketing budget.  Lastly, Period 5 accounting for 1.8% of my marketing budget.  For these last 3 years my sales fluctuated from positive and negative from the previous years, but the research costs rose and my marketing budget began to decrease.