← Back to work CASE 03 / KNOWLEDGE ARCHITECTURE

Help Center Structure & Maintenance

The knowledge architecture customers and CSMs actually use — taxonomy, voice, lifecycle, and the discipline of keeping it accurate.

Role
Owner / sole maintainer
Audience
Customers + internal teams
Output
Articles, taxonomy, governance
Cadence
Weekly triage
§ 01

What "structure" actually means

A help center isn't an article dump. It's three overlapping systems: a taxonomy that decides where things live, a voice that decides how things sound, and a lifecycle that decides what gets retired and when. Most help centers fail because one of those three is missing, not because the writing is bad.

[Iliana — describe your specific taxonomy. How do you organize the help center? By feature? By job-to-be-done? By user role? What was the logic?]

ARTIFACT
[Redacted: help center taxonomy tree, category map, or article template]
§ 02

The voice and template I established

[Iliana — what's your voice guide? Article structure? Do you use a standard pattern (e.g., goal → prerequisites → steps → troubleshooting)? Any specific writing rules you enforce?]

§ 03

The maintenance loop

The help center runs on a weekly triage tied to three signals: zero-result searches (gaps to fill), low-rated high-traffic articles (rewrites in priority order), and recurring ticket themes (new articles or product feedback). The loop is the work. The articles are the artifact.

[Iliana — add specifics about your triage process. Who's in it? How long does it take? What tools do you use to surface the signals?]

§ 04

What changed

[Iliana — outcomes. Strong candidates: article views, search success rate, ticket deflection, customer self-service rate, internal usage by CS team.]

[ ] Live articles maintained
[ ] Search success rate
[ ] Ticket deflection / CS time saved
§ 05

What I'd do differently

[Iliana — pick one honest reflection. Maybe a taxonomy decision you'd revisit, a tool choice, or a maintenance practice that didn't scale.]