Performing Constraint Profit Optimization

This chapter describes how to use the Constraint Profit Optimizer.

This chapter covers the following topics:

• About Constraint Profit Optimization
• Creating or Modifying an Optimization Profile
• Deleting a Constraint Optimization Profile
• Running a Constraint Optimization Profile

About Constraint Profit Optimization

Within the Business Modeler, the Constraint Profit Optimizer enables distributors and retailers to make the most effective use of available storage and display space throughout the supply chain. Demand forecasts for specific products are linked to space available, stock on hand, predetermined minimum and maximum levels, and profit scales to calculate the most profitable mix of products.

Note: In this example, a retail store is given as an example. However, constraint profit optimization can be applied to any stage of the supply chain, such as warehouses and distribution centers.

To achieve the best plan in a retail environment, it may be necessary to consider the following restraints, in addition to sales data and causal factors:

• Available space in store fixtures

• Available space in the store altogether

• Minimum and maximum facing required for a particular product in the store

The Constraint Profit Optimizer identifies sale opportunities, and fills the existing shelf space with products that have the highest probability of selling.

If the Optimizer discovers that a store has available or badly exploited shelf space, an alert is displayed. For example, if the minimum predetermined quantity for a product in a particular store is 10, but the Optimizer calculates its optimized inventory to be 3, the user is alerted. Similarly, an alert will be displayed if the maximum predetermined quantity for a product is 15 but the optimized inventory is calculated as 20.

The Optimizer considers the profitability of the products (not service levels). For each item, the client must give a figure on a scale of one to ten.

The system uses constraint dimensions (for example display/storage space) and a mix dimensions (for example item).

If a shelf has space which is less than can hold the maximum amounts for each product, so the Optimizer will determine the optimum stocking levels for maximum profit.

When the stock level reaches the maximum number for the most profitable product, the system then starts stocking the next profitable product.

Creating or Modifying an Optimization Profile

You can create and save any number of optimization profiles.

To create or modify an optimization profile

1. Select Parameters > Constraint Optimization.

   The Constraint Profit Optimization Wizard appears.

2. Do one of the following:

   • To create new profile, click the New Optimization Profile button.

   • To modify a profile, click the profile button.

   The New Optimization Profile - Details screen appears

3. Type a name and optional description for the profile.

4. Click Next.

   The Time and Analytical Selections screen appears.

   

5. Complete the fields as follows.

   Time Mode	Choose one of the following: Relative—specifies the time relative to the date of execution. Fixed—specifies a fixed span of time, starting with the Start Date field and spans the periods specified in the Lead field
   Start Date	Select a date for the optimization process to calculate from.
   Start After	Specify the number of time buckets after which to start the optimization relative to execution date. For example, if the number is 2 and the day of execution is today, then the optimization output will be calculated starting from 2 days from today.
   Lead	Specify the number of time buckets to use in the optimization process.
   Bounding	Choose one of the following: Bounded—Calculates the optimal stock according to both the constraint on space and product boundaries (min and max). Unbounded—Calculate optimal stock according to the constraint on space only. Both—Calculates optimal stock according to both options above.
   Include Stock	Specifies whether to include existing stock levels in the calculation.
   Statistical Model	Choose one of the following: Normal—Recommended when there is a large amount of data. Poisson—Recommended when there is a limited amount of data.
   Update Review Flag	If this option is checked, the profile updates the review flags for the combinations. If you are not using these flags, deselect this option for better performance.
   Use Prediction Status Constraint	If this option is checked, the profile considers only those combinations that have prediction_status equal to 1 (live combinations).

6. Click Next.

   The Populations screen appears.

   

   It is generally a good idea to specify the scope of the optimization by filtering the data. By doing so, you increase performance; otherwise the optimization process will run on the whole database.

7. Double-click a level in the Available Filter Levels box.

   The selected level appears in the Selected Filter Levels box.

   The bottom left side of the screen, below Available Filter Levels, now displays members that the optimization will run on.

8. Click a member in the list, and then click the right arrow button. Or double-click the member you want to remove from the optimization process.

   Business Modeler moves the selected values to the members list under the Selected Filter Levels box.

   Note: The right list cannot include more than 200 members.

9. Continue filtering the data. When you are done filtering, click Next.

   The Base Dimensions screen appears.

   

10. Complete the fields as follows.

    Maximum Constraint	Select a level or mdp_matrix series.
    Volume Unit	The unit used to measure the items.

11. Click Next.

    The Input/Output Mapping screen appears.

    

12. Here you specify the series containing the input and output data.

    Complete the fields (input series) as follows. Be sure to use only numeric series, not string or date series:

    Bounds Minimum Constraint	The lower boundary constraint
    Bounds Maximum Constraint	The upper boundary constraint.
    Profit Scale	A customer defined scale of numbers specifying profitability.
    Tolerance	Every time an item is added to the shelf, its profitability is reduced. This specifies the lowest boundary of profitability.
    Initial Stock	Stock at start of optimization process.

13. Complete the field (output series) as follows:

    Free Space

    Available display/storage space.

14. Click Finish. Or to exit without saving, click Exit.

15. Click Execute to execute the optimization profile.

Deleting a Constraint Optimization Profile

To delete a constraint profit optimization profile

1. In Business Modeler, navigate:

   Select Parameters > Constraint Optimization.

   The Constraint Profit Optimization Wizard appears.

2. Click an existing profile.

3. Click Delete.

Running a Constraint Optimization Profile

To run a constraint profit optimization profile

1. In Business Modeler, navigate:

   Select Parameters > Constraint Optimization.

   The Constraint Profit Optimization Wizard appears.

2. Click an existing profile.

3. Click Execute.

   • In the Workflow Manager, execute the optimization programmatically. Specifically, create an executable step (within a workflow) that executes the optimization program.

   • Create a workflow to call the optimization executable, passing a parameter string that is the name of the constraint optimization profile.

     Note: Depending on your options, the Constraint Profit Optimizer can affect only combinations that have a prediction status equal to 1 (live).