The following Insights post is contributed by Barbara Anderson – Former VP Operations, Product & Innovation at Les Mills US and expert JPK Group Presenter.
Forecasting is a challenge for all businesses. I have summarized 3 key steps that can give you a more robust and meaningful forecast.
Forecasts should be a continual loop of gathering, analyzing and learning. Most companies have a process for creating their forecast but what happens next? Every period (weekly, monthly) is an opportunity to review actuals to forecast and try to understand the variances.
Often, we create a forecast and leave it static because we can’t impact the quantities due to production lead times. This is a real missed opportunity. There is value in analyzing and learning from forecast variances. All of these learning should be documented and can help you improve the forecast process in the next period.
Treating the forecast as a looping lifecycle will bring new insights and improved processes that will elevate the elegance and accuracy of your forecast with each period.
Many of us forecast based off previous sales of the same or similar item. This is a good starting point, but, to make it better, we need to know what impacted these historical sales.
If you look at historical sales, what other information is associated with this data? What was happening with the company/customers/economy at that time. Did we offer a price discount that month? Did we run out of stock? Where do we store this information about things that significantly impacted sales in a prior period? For many of us, this is only stored in tribal knowledge and relies on an individual remembering events. We all know that historical sales are important enough to store but we need to treat situational knowledge with equal importance. Can we tie our marketing promotions and price changes to historical sales by event and time period? We need to start recording this level of data.
When we first create our forecasts, we make some assumptions. These assumptions should be documented and should also be reviewed on a regular basis. We may not be able to impact the amount of product coming in due to lead time, but learning which assumptions were accurate and which were way off has enormous value for future forecasts.
If our price promotion last period gave us a 20% lift in sales and we plan to do an even more effective promotion this period, we may think that we will get a 25% lift in sales. These types of assumptions need to be reviewed with cross functional teams from Marketing, Sales and Finance. We need expertise from these teams to create meaningful assumptions as they are an important aspect of our forecast.
As we move through the periods and start to look back at forecast vs actuals, we need to review, challenge and adjust our assumptions so that we can learn and continually improve the forecasting process.
Barbara Anderson is a supply chain executive with more than 15 years of executive level experience in supply chain operations, planning and forecasting. Having led several projects to design and launch forecasting systems and processes, Barbara has extensive hands on experience identifying the critical inputs to both creating and analyzing forecasts in collaborative teams. Barbara’s experience includes eight years at Mattel where she launched the Latin America Planning division in 2001. In her role as Director, she was responsible for Operations Planning for the $500 million region. Also in this role, she implemented a new forecasting process as well as spearheaded the design of an internal global forecasting system. Barbara spent eight years living in New Zealand where she co-founded an etailer fashion line before joining Les Mills to lead their clothing and equipment design and production teams. Barbara moved to back to San Francisco in 2013 where she launched the US Equipment business for Les Mills. Barbara currently lives in Marin.