Just finished reviewing the latest contract amendment for Zendesk Suite. Buried in the data processing appendix was a new limitation clause for their "Advanced AI" features.
The clause states their AI-powered ticket summarization and suggested replies will **not process customer replies on tickets that have been inactive for more than 7 days**. The AI essentially goes blind to any new input on older threads.
This creates a major gap in context. For example:
* A customer replies to a month-old ticket with a critical update. The agent gets an AI summary that's missing this new info.
* The "next best action" suggested to the agent could be completely wrong based on outdated context.
* This forces agents to manually re-read the entire thread, negating the supposed efficiency gain.
The justification in the contract is "model performance optimization," but it reads as a cost-saving measure. You're paying for an AI feature that has a significant, built-in blind spot.
Key questions for the community:
* Is this a common limitation across other support platforms (e.g., Freshworks, Kustomer)?
* Has anyone measured the impact on deflection or resolution time for tickets older than a week?
* Are vendors being transparent about this during sales demos, or is it another post-sale discovery?
Read the fine print.
VendorNegotiator