Why it matters
Freshness is trust currency customers infer from tone because UIs rarely show review timestamps. Freshness work earns that inference honestly.
Stale internal knowledge wastes payroll at scale in copilots, not only customer-facing bots.
Promotional and seasonal content needs calendar-driven freshness, not only rolling quarterly reviews. Events end; bots should stop selling them.
Authors fix content faster when they see staleness consequences on AI usage dashboards, not hidden ops metrics.
How it works
Combine calendar reviews, monitoring diffs, and test failures into freshness scores per collection.
Block or warn apps bound to stale collections past SLA. Make staleness visible to authors.
Tie seasonal reviews to calendar events, not only rolling quarterly dates.
Report freshness SLAs to leadership monthly as an operational metric.
Auto-expire promotional collections on end dates with support comms attached to release notes.
Pair freshness reviews with top-intent question lists so effort focuses on what customers ask weekly.
Example
Nintendo marks the Billing FAQ collection stale when review dates pass without owner sign-off. The chatbot shows a maintenance banner on billing intents until refresh completes.
Common mistakes
- 1Equating last indexed date with last verified date
- 2No SLA for critical policy collections
- 3Freshness tracked in spreadsheets disconnected from apps
Your source changed. Your AI should know.
Track source freshness, detect changes, and review impact before answers drift.
Explore Source Maintenance