Special Day Skip and Modes

Special days have two purposes:

• When abnormal distribution is expected in the future, you can define the days where the abnormal distribution is expected. You can then prepare a forecast to reflect the anticipated change.

  For example, if you are having a sale and you expect a 15% increase in sales. Defining that day as a special day enables you, when generating the forecast, to alter how the forecast is calculated and to factor in the expected additional distribution. Special day data is tracked by the application and can be flagged as “special.”

• When abnormal distribution has occurred in the past, you can use historical special day data to prepare a forecast when you anticipate a similar abnormal distribution. You can choose to flag certain data as special so that, when looking back, the application can apply that historical data when creating future forecasts.

  For example, if you are having an annual pre-Christmas sale for the week before Christmas. If the previous year’s pre-Christmas sales week was flagged as special days, you can use that week’s special day distribution data when forecasting for this year’s sale.

When defining a special day, you have the option to choose whether or not to skip the special day:

• When you select Yes, the data is not flagged for future use because you do not expect to have the same special day circumstances again. In most cases, choose this option if flagging special day data can impact normal forecasting.
• When you select No, the data from the special day occurrence is flagged for future forecasting. Choose this option if you want to flag the data as abnormal.

The mode option gives three ways to apply the special day distribution data:

All Days

When there is historical data available and you do not require the data to be used for specific days of the week. For example, last year the special day fell on a Monday and this year it falls on a Tuesday and there is no significant difference in expected sales distribution between the two days.

Special Days

When there is historical data available and you want to ensure they are applied using the same days of the week. For example, a 3 day sale falling Monday to Wednesday can be significantly different from a 3 day sale falling Friday to Sunday.

Percentage Adjusted

When there is no historical data available, you can estimate sales increase and decrease percentages from a regular forecast. (You must then enter a percentage value.)