Concept: Question Rotation Algorithm
The question rotation algorithm provides a steady stream of data for all your drivers on an ongoing basis, while keeping surveys short to avoid participant fatigue.
Workday Peakon is transitioning to a Flexible Listening Question model which provides greater simplicity and flexibility to our question model through a number of enhancements.
For customers who had their Workday Peakon production instances created from the 31st March 2026 onwards, you will use the flexible listening question model unless you opted out.
For customers who had their Workday Peakon production instance created before 31st March 2026, you will continue to use the existing Workday Peakon Question model until we make it possible for you to upgrade to the flexible listening question model in the 26R2 and 27R1 release cycles. Please read the below carefully to understand the impact of these changes.
Question Rotation Logic
Setting a question frequency dictates how often each employee should receive a question from each question set. When using an automated survey cadence, the survey asks active questions in each round, but not to all employees at the same time. The algorithm checks when each employee last received each question to determine if it is due again.
Manual survey frequency is an exception; it triggers all active questions of the enabled question sets at once. You cannot set question frequency for manual surveys.
How the Algorithm Prioritizes Questions
For every employee, the algorithm evaluates question importance based on:
- The time elapsed since the employee last received that specific question.
- The organization’s question frequency settings, such as monthly or quarterly.
- The total number of questions per driver and, if applicable, subdriver.
The algorithm assigns lower importance to questions and pushes them further down the queue if:
- The driver or subdriver contains additional custom questions.
- The employee has recently received questions from that driver or subdriver.
Flexible Listening Question Model Prioritization
In the flexible listening model, driver structures consist of questions with unique titles and equal weight. Because this model has no "main" questions or subdrivers, the algorithm treats all questions within a driver equally when evaluating which is due.
Classic Model Prioritization
In the classic model structure:
- The main driver question takes precedence over subdriver questions if the employee has never received it.
- If an employee submits a score that is 2 standard deviations away from the benchmark, the algorithm follows up with subdriver questions to gather more information.
Visibility Window
When a survey round ends, the visibility window (typically 1 year) shifts forward. This process incorporates new scores while gradually removing older scores from the current aggregated dashboard. Historical data remains accessible by viewing previous survey rounds, even if those scores have fallen out of the current rolling score.
Why Peakon Covers All Drivers at Once
- Covering all drivers helps surface themes that are most important to your organization quickly and objectively.
- You can access trend data within months rather than years.
- Not asking all questions at once keeps surveys short, which helps avoid participant fatigue and improves feedback quality.
Sample Sizes and Statistical Significance
Workday Peakon Employee Voice distributes driver questions by random sampling to ensure unbiased dashboard results. The system uses statistical properties of observed scores to assess when segments differ significantly from the benchmark. Peakon only assigns focus areas and strengths once results pass a significance test.
Accuracy and stability of segment scores stabilize as employees complete more rounds. In extremely small segments, wait until a majority of employees have answered before conducting comparisons. In large segments, even a small proportion of responses can provide an accurate result. Comments and topics do not require statistical validity and provide instant insight into issues.