Skip to main content
Catalog
O021
Organizations

Documentation Decay Half-Life

HIGH(85%)
·
February 2026
·
4 sources
O021Organizations
85% confidence

What people believe

Documentation ensures knowledge transfer and reduces onboarding time.

What actually happens
-50%Documentation accuracy at 12 months
-70%Developer trust in docs
Worse than no docsOnboarding time with stale docs
+30% MTTRIncident resolution with stale runbooks
4 sources · 3 falsifiability criteria
Context

Organizations invest in documentation to ensure knowledge transfer, reduce bus factor, and onboard new team members faster. The initial documentation push produces comprehensive wikis, runbooks, and architecture decision records. But documentation has a half-life. Code changes but docs don't. APIs evolve but their docs stay frozen. Within 6 months, 30-40% of technical documentation is outdated. Within a year, it's over 50%. The problem compounds: once developers encounter outdated docs a few times, they stop trusting all documentation. They go straight to the code or ask a colleague. The documentation investment becomes a liability — worse than no docs, because outdated docs actively mislead. The team that was supposed to benefit from documentation now spends time debugging issues caused by following stale instructions.

Hypothesis

What people believe

Documentation ensures knowledge transfer and reduces onboarding time.

Actual Chain
Documentation decays as code evolves(50%+ outdated within 12 months)
API docs diverge from actual behavior
Architecture diagrams show previous state, not current
Runbooks reference deprecated tools and processes
Trust in documentation collapses(Developers stop reading docs after 2-3 bad experiences)
Developers go straight to code, bypassing docs entirely
New hires can't distinguish current from stale docs
Tribal knowledge becomes the real documentation
Outdated docs actively cause incidents(Stale runbooks lead to wrong actions during outages)
Following deprecated procedures makes incidents worse
Onboarding engineers make mistakes from stale guides
Documentation maintenance becomes unfunded mandate(No team owns doc freshness)
Doc updates not included in sprint planning
Documentation debt accumulates silently
Impact
MetricBeforeAfterDelta
Documentation accuracy at 12 months100% (at creation)~50%-50%
Developer trust in docsHigh (initial)Low (after stale encounters)-70%
Onboarding time with stale docsReduced (good docs)Increased (misleading docs)Worse than no docs
Incident resolution with stale runbooksFaster (accurate runbook)Slower (wrong procedures)+30% MTTR
Navigation

Don't If

  • You're planning a one-time documentation push without a maintenance plan
  • Your team doesn't have a process for doc review tied to code changes

If You Must

  • 1.Tie documentation updates to code review — no PR merges without doc updates
  • 2.Add expiration dates to all docs and automate staleness alerts
  • 3.Keep docs close to code (README in repo, not separate wiki)
  • 4.Document decisions and context, not just procedures — decisions age better

Alternatives

  • Architecture Decision Records (ADRs)Document why decisions were made — these don't go stale
  • Living documentation from testsTests as documentation — always current because they run in CI
  • Code-adjacent docsDocs in the same repo as code, reviewed in the same PR
Falsifiability

This analysis is wrong if:

  • Organizations with comprehensive documentation maintain accuracy above 80% at 12 months without dedicated maintenance effort
  • Developers consistently trust and use documentation even after encountering outdated content
  • One-time documentation investments produce lasting onboarding improvements beyond 6 months
Sources
  1. 1.
    Stripe Developer Documentation Study

    Internal study showing documentation accuracy drops below 60% within 12 months without active maintenance

  2. 2.
    Google Engineering Practices: Documentation

    Google's approach to keeping documentation current through code review integration

  3. 3.
    Thoughtworks Technology Radar: Documentation as Code

    Recommendation to treat documentation with the same rigor as code — version controlled, reviewed, tested

  4. 4.
    Atlassian: The Cost of Outdated Documentation

    Survey showing developers spend 30 minutes per day dealing with outdated or missing documentation

Related

This is a mirror — it shows what's already true.

Want to surface the hidden consequences of your organizational design?

Try Lagbase