Value Finder requirements for strategies created outside of Next-Best-Action Designer
You can analyze a custom strategy that does not use the Next-Best-Action Designer strategy framework by running a Value Finder simulation test from the Simulation testing landing page.
Value Finder can analyze the custom strategy as long as it has the appropriate structure of eligibility rules and arbitration that contains propensities from adaptive models.
Required fields
Ensure that the strategy results include the following fields and values:
- pyStage
- This column can contain up to four values that represent the stages of the next-best-action decision funnel. You must include at least two values: Eligibility and Arbitration. You can leave the column empty when there are no actions for a customer after eligibility.
- pyPropensity
- This column must contain the propensities as obtained from the adaptive model.
Limitations
When you run Value Finder on a strategy that was created outside of the Next-Best-Action Designer framework, the option to drill down the no-action recommendations and to investigate the underlying policies, is not available. In such cases, the strategy results are isolated from the Next-Best-Action Designer framework in your system, which is why Value Finder cannot perform a complete analysis.
Previous topic Value Finder simulations Next topic Value Finder requirements for data sets