I was in the grocery store the day after Easter – a shopping cart caught my eye. It was all the leftover goodies discounted to $1.00 from $5.99. These items had not suddenly gone bad – they were just cookies and candies shaped like rabbits and Easter Eggs.
This cart of items started me thinking about other highly seasonable items – Christmas decorations that are 50-75% off the day after Christmas; Thanksgiving decorations, Halloween decorations, Fireworks after the 4th of July, the list seems to be endless.
Some items such as ice cream and sports drinks have high seasonal spikes. Ice cream manufacturers will rent additional cold storage warehouse space from third parties in order to prebuild months in advance enough ice cream to meet the summer spike.
But what about items that have a narrow window during which time they can be sold such as seasonal decorations and tax software. These items essentially have little or no perceived value after their usage period has passed.
These commodity shelf life practices goes to show the importance of having an accurate forecast. Often in a company, the forecast accuracy is a common water cooler joke and schedulers laugh at how inaccurate it always seems to be. Companies will typically account for this inaccuracy by increasing safety stock to ensure high customer service levels, but this is not practical for highly seasonal items. The cost of shipping the item back to a warehouse and storing for another year is more than just letting the item go at a fire sale price.
Improper forecasting of these highly seasonal demand items can cause whiplashing of your entire supply chain and unexpected fees for expedited freight.
Forecasting is both a science and an art, and there are some simple rules to keep in mind:
Forecasting is a Journey
The goal is a process that generates a reasonable forecast, not perfection. Don’t expect your forecast to be 100% accurate on the first try. As you gather more data, you’ll refine your approach and see your forecasts become more precise. The important thing is to start! You can plan and adjust as you go, whether you’re looking at short term highly seasonal items or taking a long-term view
Gather Input Across Your Supply Chain
It is important as you create your forecast, to gather different input from every function of your organization – start with sales and move through the supply chain to operations. This process will not only allow you to anticipate demand but also help you understand how much supply-side capacity you need to meet the forecast. This process can ultimately help you develop a good capacity constrained production schedule.
History Does Not Necessarily Repeat Itself
Just because your business is seasonal doesn’t mean it’s going to see the same patterns year after year.
Maybe you had a lot of orders last January, but that doesn’t mean your orders won’t come a month early or late next year. External factors often play into actual demand from year to year. Keep evaluating the actual results you experience against your forecasts to refine them further and make sure your production is throttled up or down as needed.
The Same Forecasting Model Does Not Work for Every Business
Forecasting is a statistical science. There are many different forecasting models and tools. Forecasting can quickly get very technical and complicated. Don’t get caught up in technicalities at the beginning. The most important thing is to pick an approach and start.
Every forecast will always have a component of error, regardless of whether your business is stable, cyclical or seasonal—the key is to accept this and continue to improve your forecasting to minimize this error.
Constantly Improve and Adjust Your Forecasting Methods
Developing a good forecasting method is important, but it must constantly be evaluated and adjusted. By continuously refining your forecasting and apply what you learn to your operations, you can achieve a better balance between supply and demand.
A good forecasting method is like owning a car – it requires proper care and upkeep to operate efficiently.