- Adoption gap on required filtering: Our POC shared a major insight from a recent check-in with one of her top power users. Even experienced users are no longer consistently setting the Philips Solution custom field before using AI Assistant or Recommendations, despite the enablement you have already delivered (quick reference guides, internal posts, emails, and 1:1 coaching).
- Impact on answer quality and risk: When the custom field is not set, users can receive irrelevant answers. Our POC gave a concrete example where AI-generated content surfaced the wrong type of response for an RFP, reinforcing that this is a material workflow and quality risk.
Product Enhancement Request:
- Product enhancement idea (mandatory prompt): We aligned that a forced prompt (or required selection) before AI/Recommendations begins could materially improve outcomes, especially if it is designed as opt-out rather than opt-in.
- Filtering philosophy: You clarified that the single most critical control is the Solution custom field. Collections generally “take care of themselves,” and your library structure is intentionally designed so setting the solution is the main “price of admission.”
Today, our POC will start with Announcements to her entire user base to see if this might force the process of checking filters before running AI Assistant, Recommendations, and AI Draft.
