In this article
The What-if widget predicts the impact of enhancing customer experience by adjusting averages of key drivers. Key driver models are dynamic and personalized.
The what-if widget utilizes an existing Key driver widget model and employs a regression algorithm, enabling report viewers to visualize the predicted impact on a specified KPI by adjusting the averages of the drivers included in the key driver.
This enables the user to comprehend the predicted impact of enhancing crucial aspects of the customer experience and identify which enhancements can lead to the greatest change in their key metric by conducting scenarios and comprehending the probable outcome.
Note
The widget is not capable of predicting based on aggregated metrics, such as the top 2% of a question or Net Promoter Score (NPS), due to the limitations of its regression algorithm. In other words, it doesn't factor in how the probabilities of outcomes can vary depending on the value of a given variable. Thus, it's unable to provide meaningful insights about proportion-based measures.
Example of the What-if widget in use
This allows the user to understand the predicted impact of improving key areas of the customer experience and see which improvements can drive the biggest change on their key metric by running scenarios and comprehending the likely outcome.
Please keep in mind that as Key driver models are designed to be dynamic, they consider any level of personalization (such as hierarchy) and filtering. As a result, a valid model for one user viewing a specific data profile may differ from another user viewing a different data segment. Guard rails are in place to account for these varying impacts based on data segmentation.