Suggested articles based on URL Context Sensitive content | Community
Skip to main content
Submitted

Suggested articles based on URL Context Sensitive content

Related products:Knowledge & Help Center
  • July 18, 2025
  • 0 replies
  • 37 views

Forum|alt.badge.img

The problem: The Suggested Articles widget in Messenger shows AI-generated suggestions, but they're often not relevant to the specific page the user is currently viewing. There's no way to target suggested articles by URL — we can't assign specific Help Center articles to specific pages of our app. Users on a specialized screen get generic suggestions, give up on self-service, and open a conversation instead.

Example: A user is on our Special Orders dashboard (www.ourapp.com/specialorders). We have a Help Center article written specifically for that feature — but we can't make it appear there. Meanwhile the AI surfaces loosely related articles that don't answer the question the user almost certainly has on that page.

The request: Let us assign Help Center articles to specific URLs and URL patterns, with wildcard support for dynamic segments — e.g., /account/*/settings should match every user's account settings page regardless of the ID in the URL.

Intercom already supports current-page URL targeting for outbound messages, tours, and checklists — this request is essentially extending that same, proven targeting capability to Suggested Articles.

Supporting details:

1. Priority: Manually assigned articles should appear above AI-generated suggestions, so curated content always leads.

2. Hygiene: If an assigned article is archived or deleted, it should automatically drop off the suggestion list — no broken or stale links.

3. Reporting: Click/engagement data on assigned articles, so we can see which page-level suggestions actually deflect conversations and optimize accordingly.

The payoff: more relevant self-service at the exact moment of need, fewer unnecessary conversations reaching the team, and better use of the Help Center content we've already written. This would also strengthen Fin — knowing which articles are most relevant to the page a user is on gives Fin better context to resolve questions on the spot, improving resolution rates for everyone. Thanks for considering!