A Win Rate model needs to fetch data from either a Data Source or a Datamart. The source should contain at least 1 years of quote history including quote status, each record should be a quote item.
following columns:
|
Field |
Required? |
Comment |
|---|---|---|
|
Product |
Yes |
Typically Product ID or SKU |
|
Customer |
Yes |
Typically Customer ID |
|
Quote |
Yes |
Typically Quote ID or Number |
|
Quote Date |
Yes |
|
|
Revenue |
Yes |
Extended to the quantity. Will be used to compute the price |
|
Quantity |
Yes |
|
|
Profit |
Yes |
Extended to the quantity. Will be used to compute the Margin Rate and Costs. |
|
Revenue at List Price |
No |
Recommended, mandatory if you want to compute the Discount Rate. List Price * Quantity. |
|
Win/Lost |
Yes |
It is the quote status. Then you set the values defining a win and a loss, which will be required in the interface.
|
|
Categorical features |
Yes |
All features as categories (text values) may have an impact on the win rate, from product category, customer segment, product competitiveness, region… Those features are crucial for the accuracy of the model. |
|
Numerical features |
Yes |
All features as number may have an impact on the win rate, like competition positioning index, size, horsepower… Those features are crucial for the accuracy of the model. |
Quote status is crucial for the model so please be mindful of defining properly what a win and a loss are. It makes sense to ignore quotes like variant quotes, revised quotes, not approved…