Startup Curve

Startup Curve

What are you going to learn?
  • How to create a startup curve
  • How to edit an existing curve
  • How to delete a startup curve

The startup curve allows you to model new items, without sales history, after items similar to them.

You can locate the startup curve page by clicking the "Forecasting" drop-down menu and selecting "Startup Curve."



You will then be directed to a screen that looks like this:

1. SKU - Select the SKU that you wish to apply a startup curve to in your filter sidebar.

2. Apply - Click apply after you have selected the SKU you wish to add a curve to.

3. View Existing Curves - View and edit any curves that you have already created. (This should not appear if you have not created a startup curve before.)

You can select that SKU by locating it in the categories and filters section under "SKU" and then clicking "Apply."


Clicking "Apply" will create two new boxes allowing you to apply a SKU to model your new SKU after. In the graphic below you'll see that we have chosen 1100-101) to be modeled after 1000-101.


1. SKU used to model after - The new SKU, 1100-101, will be modeled after this SKU, 1000-101.

2. New SKU - The SKU that we are applying the curve to.

3. Apply - This apply button is used to select the SKU to be modeled after (1000-101).

Edit Settings

Once you've selected your new SKU and the SKU you want to model it after, click the grey right arrow at the bottom of the page. You will then be brought to a page showing a graph representing the source history of the old SKU and the logarithmic regression of the new SKU.

This screen is divided into 5 different sections: Seasonality Ratios, Source History Graph, Source History, Startup Curve Dates, and Logarithmic Equation. These sections are explained in further detail below.

1. Seasonality Ratios - Shows the ratio of variance between the new and old SKU’s projections.

2. Logarithmic Equation - Shows the correlation with the sales history and the equation used to generate it.

3. Source History Graph - Shows a visual representation of the old SKUs sales history and the log regression.

4. Source History Calculations - You can alter the source history either using the oldest,most recent, or custom sales period. The oldest will represent the old SKUs sales history from the first month it was launched, until 12 months from that month.

5. Next/Back Arrows - Help you navigate back and forth between screens.

Settings Confirmation

1. Seasonality Ratios - You will again see the seasonality ratios but you will now be able to alter them by clicking in their respective boxes.

2. Forecast Points In a Chart - Shows the graph visually representing the data in the table below.

3. Forecast Points In a Table - Represents the exact values of the points you see here in the graph. You can also see their values by hovering over them.

4. Start Date - The date that the curve will begin. For example, if the month is currently december, but you don’t want the new SKU to forecast until February when you release it to the public, you can set February 2020 as the start date.

5. Expires After - Allows you to determine when standard forecasting will begin in a location. For example, if we select 18 months, once these locations reach 18 months’ worth of sales, it will automatically stop using the curve and begin running a standard forecast.

6. Summary Table - Each of the locations you have imported with the new SKU and their ratios. These ratios determine by what the forecast of that location will be scaled. So If you believe that Houston will be twice as successful as Baltimore, select a 1 for Baltimore and a 2 For Houston.

7. Logarithmic Equation - Allows you to alter the equation being used to forecast. You can change the number in brackets by clicking in the box and altering the number. The number in brackets represents the starting month of data used. For example, if the old SKU began selling in January, putting a 1 in the brackets would start with January 2017. Putting a 2 would start in February 2017,a 3 would start in March 2017, so on and so forth. The curve correlation represents the relationship between the log regression and sales history.

8. Multiply Factor - Sets what the equation is multiplied by. If you believe the new SKU will be twice as successful as the old SKU starting out, you could put a 2 here.

9. Log Expires After - Allows you to set the amount of months of sales until the new SKU generates a new log based off of its sales. So if we select 3 months here, and you choose to start a curve on February 1st, the new SKU will generate a new log on May 1st based off of the SKUs sales in February, March, and April.

10. Aggregate Expires After - Similar to the log, but focuses on each individual location rather than the SKU itself.

11. Notes - You may explain why you may have set certain settings the way you did.

12. Curve Name - Name the curve here, which will help others in your company locate the curve and make adjustments as necessary.

13. Apply Forecast - Applies the curve to the new SKU and redirects you to the forecasting screen.

14. Back - Move back to previous screens in order to alter settings.

chat with us