Optimization Model Logic

The Optimization calculation logic is the core of the optimization model because the whole model behavior is defined in the logic: most of the user inputs, calculations, policy records, etc. For the standard Segmentation-Optimization model it is not necessary to implement the logic from scratch because we provide a customizable reference implementation.

In general, Optimization is intended to broadly meet three main use cases:

1. Standard segmentation and optimization using a normalizable transaction-level measure such as net margin (profit) or discount. These logics are available via our Customer Solutions.

2. Same as 1 but with minor customizations such as label changes, custom score calculation, etc. In this case, the reference logics are tweaked in a few elements: Data, SegmentCalculation, etc. There is no predefined recipe, as we want to take a full advantage of the configuration capabilities of our platform, but generally the configuration effort is expected to be less than a week.

3. Custom implementation, where the customer's requirements clearly do not fit our standard distribution-based segmentation and optimization model. Expect several weeks of configuration effort.

Most implementations will fall in categories 1 and 2. We will elaborate somewhat on makeup of an Optimization logic, as it will also help in category 2 efforts, but category 3 is deemed to be a rather exceptional, very advanced use case requiring engagement of Customer Solutions, including Pricing Science.

Further reading: