Text Analytics

Typically, Medallia Text Analytics is only recommended for confidential Employee Engagement programs, for users with Insights/Admin roles (other may not have sufficient data to generate insights).

Consider these situations when implementing Text Analytics:

  • In the Topic Deep Dive and Theme Deep Dive modules, create separate Responses Feed modules for each survey question to avoid seeing multiple responses from the same employee together.

  • Do not use TA Topics/Themes Comparison module, since it has an out-of-the-box Responses Feed with all questions and does not allow to split the output into multiple Responses feed modules. Instead, use TA Pivot Table.

  • The TA Topics/Themes Explorer module does not support hiding time fields. Ensure the Reporting date field is set to the Wave Date so that the Response date is not visible in reporting.

  • These Text Analytics modules over-mask data due to Minimum sample size restrictions:

    • TA Topics/Themes Snapshot and TA Topics/Themes Trends Over Time apply minimum sample size at the cell level, not a a row level. In this screenshot, % Positive is displayed but % Negative is masked because the topics have fewer comments with negative sentiment than the minimum sample size.

      Topics/Themes Snapshot does show a number for the % Negative metric.

    • The TA Topics/Themes Explorer bubble visualization does not show colors around a bubble unless there is sufficient sample of positive, negative, and neutral (if enabled) comments.

      Topics/Themes Explorer in a bubble visualization.