Forecasting is one of the core features of Inventory Planner. Forecasted sales drive your recommendations for Replenishment.
Forecast settings can set based on multiple forecast methodologies:
- Recent Sales and Trends
- Last Sales
- Top Down
- Top Down Seasonal
Recent Sales and Trends
This method uses recent sales and stockouts during the Sales Period for Forecast when generating a forecast. Inventory Planner calculates sales velocity and trend based customer orders placed during this time frame.
Note: Holiday sales such as Black Friday, Cyber Monday and the week before Christmas are excluded with this method so that unusually high sales do not skew forecasting calculations.
The Sales Period for Forecast and Trend can be adjusted by Variant by clicking the "i" icon under the 'Details' column, clicking the 'Forecast Settings' tab, and enabling 'Use Custom Forecasting Settings'.
This forecasting method is best for non-seasonal products. It is based on the average sales for the last "X" number of days, where you set the number of days.
If the setting is enabled to "adjust forecast using stockouts," the forecast will only look at periods when the product had available stock to calculate the sales velocity. Here is an example of how the algorithm works if "15" is entered for last "X" days:
- 1. Look at sales and stock over the last 15 days.
- 2. If the product went out of stock during the last 15 days, we look at the past 15 days when there was available stock to calculate the sales velocity. Note: sales during stockout periods are ignored.
- 3. If the product had no stock in the last 15 days, we use the last sale date when it was available, then look at sales and stock from the prior 15 days to calculate the sales velocity.
- New accounts: Inventory Planner tracks stock history starting with the day you connect your platform. Since we cannot see past stock, we will look at the 'Last Sold At' date then add sales from the prior X days to calculate sales velocity.
Exclusions: the following are ignored using this forecast method;
- Manually inputted stockouts.
- Manually adjusted sales history.
- Wholesale orders.
This forecast method refers to the same months in prior years as the sales reference period. This is beneficial for holiday and seasonally driven items with 12+ months of history. Sales Velocity is not a consideration in this case. Trend is applied by evaluating sales over the last 12 months vs. the prior 12 months.For example, sales from January 2018 and January 2019 impact the forecasted needs for January 2020. If today is December 31, 2019 - the trend will be based on sales between January - December 2019 vs. January - December 2018.
This setting computes a seasonal forecast at the Category level*, then distributes forecasted sales to the Variants. The last 2 months of sales are used to determine each variant's % contribution to the total unit sales of the category. For example, if Variant A contributed 5% of the sales units to Category Z during the last 2 months, then Variant A will receive 5% of the category's forecasted sales in upcoming months.
*Note: if a category has less than 12 months of history, forecasts are first computed at the store-wide level then distributed to the Variants.
Top Down Seasonal
This setting computes a seasonal forecast at the Category level*, then distributes forecasted sales to the Variants. When an item has 12+ months of sales history, the same time period in prior years is used to determine each variant's % contribution to the total unit sales of the category. For items with less than 12 months of sales history, we default to the top down, non-seasonal method.
For example, if Variant A contributed 5% of the sales units to Category Z last December, then Variant A will receive 5% of the category's forecasted sales for next December.
ACCOUNT-WIDE FORECAST SETTINGS
The default Sales Period for Forecast is set for all items at the top right of the Replenishment screen. Customer orders placed during this time will be used to calculate the forecasted demand.
For non-seasonal forecasts, Inventory Planner calculates the sales velocity (average sales per day excluding the number of days out-of-stock) during this time frame, and the trend based on the input for 'Trend Months' under Account Settings.
For seasonal forecasts, Inventory Planner refers to the same months in prior years as the sales reference period. This is beneficial for holiday and seasonally driven items with 12+ months of history.
Account-wide forecast settings can be found by going to Account > Settings:
Use Top-Down Forecasting: This setting computes a Seasonal forecast at the Category level - or at the Storewide level if a category has less than 12 months of history - then distributes forecasted sales to Variants using the logic explained in the "Top Down" and "Top Down Seasonal" sections below (under Custom Forecast Settings).
Enable forecasting for wholesale customers: This setting allows the ability to separate wholesale orders from regular, replenishable demand. Replenishment recommendations will add wholesale needs on top of regular demand. Learn more about wholesale planning here.
Adjust forecast using out-of-stock information: Inventory Planner automatically detects when a product is out of stock starting from the moment you connect your store to Inventory Planner. This data helps to estimate the demand correctly.
For example, if you sell on average 10 units of Variant A per month and the product is out of stock for 1/2 of a month - the real demand for it is 20 units. Without taking stockouts into account, the forecasted demand could be too low.
By default, your account is set to take into account stockouts. If your store allows overselling (continuing to sell an item even when it is out of stock), then you should change this setting to disregard out-of-stock information.
Products are seasonal by default: This setting enables Seasonality for all Variants. This forecast method refers to sales during the same months in prior years as a reference for forecasting. This is beneficial for holiday and seasonally driven items as long as there is at least 12 months of history. Trend is applied in this case by evaluating sales over the last 12 months vs. the prior 12 months.
For example, sales from January 2018 and January 2019 impact the forecasted needs for January 2020. If today is December 31, 2019 - the trend will be based on sales between January - December 2019 vs. January - December 2018.
Produced bundles by default: This setting considers all bundles to be produced bundles. If all bundles in the account are assembled prior to fulfillment (that is, the stock of bundles is separate from stock for each of its components) then this setting should be enabled. If bundles are assembled at the time of fulfillment, then do not check this box. Learn more about bundles and packs here.
Trend Months: This field sets the number of consecutive periods with a pattern to define a trend.
In a non-seasonal forecast, if 'Trend Months' is set to 2 with increased sales for an item from March to April and again from April to May, a growth trend will be incorporated when forecasting June.
Default Replenishment: This field sets default replenishment recommendations for items with zero stock and no sales history.
CUSTOM FORECAST SETTINGS
The Forecasting Method can be adjusted by Variant on the Replenishment page by clicking the "i" icon under the 'Details' column, clicking the 'Forecast Settings' tab, and enabling 'Use Custom Forecasting Settings'. Alternately, forecast settings can be updated by selecting variants using the checkboxes on the left of the Replenishment screen, then clicking on Bulk Actions > Set Forecast Settings.
ADJUST FORECAST USING STOCKOUTS
Adjust forecast using out-of-stock information can also be set by Variant by clicking the "i" icon under the 'Details' column, clicking the 'Forecast Settings' tab, and enabling 'adjust forecast using stockouts'.
Past Stockouts can be adjusted manually under the 'Edit Forecast' tab, or under Details > Edit Forecast. Adjustments can also be imported through a CSV. More information on adjusting stockouts can be found here.
Select multiple items using the boxes on the left to display "Bulk Actions" at the bottom of the screen. Click Bulk Actions > Set Custom Forecast Settings.
Note: If you need to make additional changes to your forecast, read our article about the Edit Forecast functionality.