Forecasting & Replenishment

What do New Year’s resolutions have in common with Demand Planning?

Whether you see them through or toss them out the window at the first opportunity, New Year’s resolutions are often based on a behavior or situation from the past that you’d like to change in the future. In a similar vein, if you want your demand planning efforts to succeed, the first thing you need to do is be sure you’ve got accurate historical usage data.

Don’t set yourself up for failure

Too often demand planning systems fail because little attention is paid to the accuracy of historical usage. After all, if you don’t have good usage data to start with, you won’t have a good forecast.

By usage data, we don’t mean what’s sold, or the total number of requests, or shipments. In this context, usage is a special history derived from actual sales, but adjusted based on when the customer actually wanted the product. It is a fine distinction, but rather than indicating what has moved through your warehouse, historical usage accurately reflects what the customer wanted (not necessarily what was shipped).

If you disregard this subtle distinction, you could end up putting a lot of time into assigning formulas to usage data that is not consistent, and you’ll never get the forecast accuracy you desire.

However, a forecasting system with a good best-fit formula selection process (like our Demand Planning system) will perform very well when it works from consistent historical usage data. By putting in place a demand planning system that helps ensure historical data is accurate and consistent, companies can optimize their purchasing in order to effectively manage their largest and most costly asset – inventory.

In a nutshell, the problem with many forecasts involves the data used to forecast future demand, not the actual forecast formula itself.

Sales promotions — an example

Achieving historical accuracy can be very challenging in a number of areas. Sales promotions, for example are typically viewed as a good thing. Many times they are used to reduce an overstocking situation. At other times promotions may be done in cooperation with a vendor who is lowering your purchase price, and you wish to pass this along to your customers.

So the increase in sales is great, but what does it do for your historical usage? Will your forecast be artificially inflated in the future?

Here’s what we recommend:

• A sales promotion should be set up in advance of the vendor’s lead time in case you need to purchase inventory to cover it.
• The promotion should include the date and the anticipated increase in sales, stated in quantities. This information will then be used to increase the forecast for the corresponding period of time.
• After the promotion is over, an analysis will reveal how much the actual sales did increase, and this increase should be adjusted out of historical usage in the same periods.
• While a sales promotion increases the forecast, it should also be used to remove the effects of the promotion on sales history afterward.

***********************************************************************

If your New Year’s resolutions include reducing inventory levels, increasing profits, and improving customer service levels, we can help! Let us know if you have questions, and consider joining us at our upcoming Forecasting & Replenishment Forum, May 8-10 in Scottsdale, AZ, where we’ll be discussing the importance of obtaining accurate historical usage and much, much more!

#### ####