← Previous experience

Learning experience 04

Keep the system working

Meridian solved the discoverability problem with Source. The next challenge was operational drift. Product launches, regional exceptions, and process changes were happening faster than the knowledge system could evolve. This is the governance model I designed to keep it current.

1 of 4
01

Monitor operational signals continuously

Source is reviewed through operational behavior — not documentation volume. Repeated questions, failed searches, and escalation mismatches are treated as indicators that the system no longer reflects real workflows.

Signals tracked

Repeated internal searches on the same topic — signals a content gap or a clarity failure
Incorrect escalation rates across regions — signals a decision tree that no longer reflects current process
Manager-dependent questions reviewed weekly — signals where Source is not being trusted or found

These signals don't require surveys or manual audits. They're already happening. The governance system makes them visible and actionable.

2 of 4
02

Use AI to identify what needs updating

AI connects patterns across three data sources to identify where Source is failing operationally — before it becomes a performance problem.

Data sources

Source 1
Support tickets
Source 2
Search behavior
Source 3
GTM launches

AI flags

Articles linked to increased escalations after product or policy changes
Search queries with low resolution — users searched but didn't find what they needed
Regional inconsistencies in process execution or terminology

Example: Meridian launches a new escalation path for enterprise clients. Within 72 hours, Source detects repeated searches for outdated escalation criteria in EMEA and flags the exact decision tree requiring revision.

3 of 4
03

Tie governance directly to launches

Product and GTM launches automatically trigger operational review tasks inside Source. The system evolves at the pace of the business — not on a separate quarterly cycle that's always one launch behind.

Launch-triggered reviews

Onboarding flows impacted by the launch are flagged for immediate review
Decision trees and escalation logic are updated before the change goes live
Role-based views are audited for regional inconsistencies introduced by the launch

Enablement owns the governance model and structure. Functional teams own content accuracy. Clear ownership at both levels prevents drift without creating bottlenecks.

4 of 4
04

Define ownership before scale creates friction

Every article inside Source has a clear owner, review cadence, and operational purpose. At scale, undocumented ownership is how good systems quietly fail.

Governance model

Quarterly reviews archive outdated or duplicated content — volume is not a quality signal
Regional exceptions are documented and approved centrally before being published
All structural changes include documented rationale — the system stays transferable
Done

How we know it's still working

Source became an operational layer, not a documentation system

Measured at 6 months post-launch across all regions. These aren't training metrics — they're business outcomes.

Time to productivity
14w → 8w
−43% across all regions. Fastest improvement in APAC.
Regional consistency
61% → 89%
Spot-check audits across 10 countries, 6 months post-launch.
Incorrect escalations
−38%
Decision trees reduced escalation mismatches globally.

A knowledge system should reduce operational friction — not become another system employees need to manage. The governance model is what makes that possible at scale.