Hi Team, we have a couple of suggestions for the generative search feature in the Knowledge product that we hope will improve the experience for admins.
- {{generative_answers}} placement validation
When using a custom Zendesk theme, enabling Quick Answers requires manually inserting
{{generative_answers}}intosearch_results.hbs. While the system correctly flags some incompatible placements with an error, it does not consistently catch each placement — allowing the template to be saved and published with no warning while feature silently stops working. This primarily affects admins responsible for theme configuration.The system should consistently validate all placements of
{{generative_answers}}insearch_results.hbsand prevent saving or publishing when an incompatible location is detected. -
Generative search usage
Usage tracking for Generative Search is currently available in Admin Center under Account > Usage > Summary. However, the data is aggregated at the account level and lacks the granularity needed for multi-brand instances. This affects admins and stakeholders who need per-brand visibility to make informed decisions about feature adoption, capacity and performance.
Without brand-level usage breakdowns, it is not possible to understand how different brands within a multi-brand instance are consuming the feature, making it harder to attribute usage, optimise content, and plan for scaling.
The ideal solution would be to add a detailed usage table within the Generative Search section showing monthly consumption broken down by brand, covering at least the trailing 12 months broken down by month — enabling admins to compare brand performance, and make data-driven decisions over time.
Looking forward to see these improvements in a future update.
Thank you.
