What matters
- Credibility and freshness are separate dimensions.
- Different fields need different review intervals.
- Expired evidence should trigger a review state, not automatic deletion.
- Time-sensitive commercial decisions should fail closed when evidence expires.
Static pages create a false sense of permanence
A fee page, market list or product term can be correct on publication day and wrong months later. Search engines may continue serving the page long after the operational fact has changed.
A research system therefore needs to model time as part of evidence.
Not every fact ages at the same speed
A founding year may be stable. A fee schedule, supported country, bonus or derivatives rule may change frequently. Review cadence should match the volatility of the field instead of applying one site-wide freshness badge.
Records can store source date, last verification and next-review date independently.
Expiry is not the same as false
When evidence expires, the correct state is 'needs review', not automatically 'incorrect'. Historical evidence can remain useful for audit and editorial context.
Public UI can suppress stale time-sensitive values while keeping the source in the internal record.
Freshness also protects partner relationships
Partner campaigns, tracking URLs and market permissions often have explicit validity windows. Expiring those rules automatically prevents old promotions from lingering after commercial terms change.
The result is a platform that scales with more partners without depending on memory or manual cleanup.
Primary reading
Official source material used for background and risk framing. Platform-specific claims should be verified against current operator documentation and local rules.