We built a comprehensive QM evaluation form with 45 questions organized into 6 sections. Several questions use conditional visibility - for example, Question 12 (“Did the agent offer a callback?”) only appears if Question 11 (“Was the customer placed on hold?”) is answered “Yes.”
The conditional logic works correctly on initial evaluation. If the evaluator answers “No” to Question 11, Question 12 remains hidden. If they answer “Yes,” Question 12 appears.
The bug: if the evaluator initially answers “Yes” to Question 11, answers Question 12 with “No,” and then goes back and changes Question 11 to “No,” Question 12 remains visible with the previously entered answer. It does not re-hide.
This is causing scoring inconsistencies because the answer to Question 12 is still counted in the total score even though the parent condition is no longer met. The evaluator does not realize they need to manually clear Question 12.
We have 15 conditional question chains in the form and this behavior applies to all of them. The form was created in October 2024 and has been in production since November.
Is this a known bug with the evaluation form conditional visibility, or is there a configuration option to force child questions to reset when the parent answer changes?
This is a known limitation in the QM evaluation form engine, not a configuration issue. The conditional visibility logic is evaluated only once when the question first renders. It does not re-evaluate when the parent answer changes after the child question has already been displayed.
Genesys has acknowledged this as a bug (tracked internally, no public ID available) but it has not been prioritized for a fix because the workaround is straightforward.
The workaround is to restructure your evaluation form to use Sections with conditional visibility instead of individual question-level conditions. When you set a condition on an entire Section, the section visibility IS re-evaluated when the parent answer changes, and all questions within the hidden section are excluded from scoring.
So instead of:
- Q11: Was customer placed on hold? (always visible)
- Q12: Did agent offer callback? (conditional on Q11=Yes)
Restructure as:
- Section A: Hold Management (always visible)
- Q11: Was customer placed on hold?
- Section B: Callback Offer (conditional on Q11=Yes)
- Q12: Did agent offer callback?
When the evaluator changes Q11 from Yes to No, Section B hides and Q12’s score is excluded from the total.
Oh wow, we just ran into this exact same issue last week! The section-based approach works perfectly. Just wanted to add that when you restructure the form, you do NOT lose historical evaluation data from the old form version.
Genesys Cloud keeps all evaluation data associated with the form version that was active when the evaluation was completed. When you publish a new form version with the restructured sections, it only applies to future evaluations. Past evaluations retain their original form layout and scoring.
So it is safe to restructure the form mid-quarter without affecting your historical QM reporting.
Section-based conditions are the fix. If you want to also programmatically validate that no orphaned answers exist in completed evaluations, you can use the Quality API to audit:
GET /api/v2/quality/evaluations/query
Iterate through the returned evaluations and check if any answered questions have a parent condition that evaluates to false. Flag those evaluations for re-scoring.
We built a nightly audit script that catches these orphaned-answer cases and sends a notification to the QM lead with a list of evaluations that need manual review. It caught 23 mis-scored evaluations in the first month.