# Seasonality

### Some products, such as running shoes, are more popular in summer than in winter. Products such as heaters sell best during the cold winter months. These products are seasonal and this situation calls seasonality.

To predict the future consumption of these types of items correctly, you must take that seasonal sales nature into account. The best way to do this is to use sales data of products that are comparable in terms of seasonality (if you do not have data on the particular product you want to analyze).

Let’s presume that the sales of the product in 2010 were as follows:

Our task here is to get adjustment factors that we can apply to forecasting data for 2011(assuming that the character of seasonality of the product remains constant). To do this, we calculate the average monthly consumption for the past year relative to all of 2010. For each month in 2010 we calculate the variation factor from the average value by using the following formula:

**Monthly consumption / Annual consumption = Seasonal factor**

According to this table, we can estimate that the annual average for April 2011 is equal to: 160/0.72, which comes to 222 pieces. We can now estimate future consumption for 2011. The correction factor for May is: 222 * 1.44, which comes to 320 pcs. Similarly, we can make adjustments to the average daily consumption for each month of the year.

The algorithm can be further adjusted by using the annual average estimate for the previous period; i.e. calculating the annual average estimate for each previous month, obtaining an average estimate of the obtained values, and then calculating the estimate of sales for the next month using these numbers. The chart below provides the resulting estimates:

Here, the estimated annual average for May is calculated as follows: You add the estimated annual average for the previous four months (277 + 250 + 234 + 222) and divide by 4 to come to 246 for May.

To estimated May sales you multiply that 246 by 1.44 (1.44 is May’s seasonable value) to come to 354.3. This is an accurate estimate of future consumption that takes seasonality into account.

Mycroft Assistant allows you to choose future consumption rates that take seasonality into account using the second algorithm. The solution allows to specify your own sales seasonality data for each group of products or product items:

The factors of average deviation in a given month relative to average annual consumption are specified at the bottom of the chart. As you can see, they and are set from 0 and up, with a default value of 1. The default value of 1 indicates that monthly consumption is equal to the annual average. You would insert a value of 0.5 to indicate that a month’s consumption is half that of the annual average; a 2 would indicate that the consumption for the month i is twice the annual average (and etc.)

If you do not want to manually indicate seasonality, this value can be automatically calculated by the solution based on sales figures of the previous year. To do this, you would simply click on the “Calculate” button. If there was no sales input for the product for the previous year, the solution will notify you, and you can specify a specific example product that Mycroft Assistant will then use too calculate the seasonality of this product.

Know more:

— Demand forecasting – information about it you can find HERE

— Sales analysis – information about it you can find HERE ;

— Planning – information about it you can find HERE and HERE ;

If you want to get started with the Mycroft Assistant – where we have implemented everything necessary for you to work effectively, simply sign up HERE and follow the provided instructions.