How often should I update content for AI freshness? My 2-year-old guides are invisible

AI systems have dramatically changed the freshness game.

The new reality: What used to stay visible for 24-36 months now needs updates every 6-9 months. AI systems heavily weight recency in citation decisions. A newer, shorter article often beats an older, comprehensive guide.

Why this happens: AI systems train on and retrieve from the web. They associate fresh content with accuracy. When a user asks a question, AI wants to cite current information. Old content signals potentially outdated information.

The freshness hierarchy:

Our findings: Recently updated content appears 4.3x more often in AI answers. The sweet spot is updates within the last 90 days. After 6 months without updates, citation rates drop significantly.

This is a fundamental shift from SEO thinking where evergreen meant “set and forget.”

What counts as a real update vs. cosmetic change:

NOT substantive (don’t bother):

• Changing a few words
• Fixing typos
• Updating the date alone
• Minor formatting changes
• Adding a sentence or two

Minimally substantive:

• Adding 1-2 new paragraphs (200-300 words)
• Updating 3-5 data points
• Refreshing 1-2 examples
• Minor FAQ additions

Substantive (what you should aim for):

• Adding new section (500+ words)
• Updating all statistics to last 12 months
• Refreshing all screenshots/visuals
• Adding 3+ new FAQ questions
• Significant rewrites for clarity

Our update checklist:

For each refresh:

• Add at least 500 words of new content
• Update all statistics to current data
• Refresh examples to 2025 context
• Update all screenshots/visuals
• Add 2-3 new FAQ questions
• Revise intro to acknowledge recent developments
• Update meta description
• Set accurate modified date

The test: Would a returning visitor notice meaningful improvements? If yes, it’s substantive. If no, don’t bother updating the date.

You can’t update everything. Here’s how to prioritize:

Priority Matrix:

Tier 1 identification (update every 60-90 days):

• Top 10 converting pages
• Top 10 traffic pages
• Pages currently cited in AI answers
• Core topic pillars

Tier 2 identification (update every 6 months):

• Supporting content for Tier 1 topics
• Category and hub pages
• Content with moderate traffic
• Pages that used to rank well

Tier 3 identification (annual review):

• Foundational definitions
• Stable reference content
• Low-traffic but necessary pages
• Content on topics that rarely change

For your 150 articles:

• Tier 1: ~15-20 pages (update quarterly)
• Tier 2: ~40-50 pages (update biannually)
• Tier 3: ~80-90 pages (annual audit)

This is manageable vs. trying to update everything.

Our 90-day content refresh workflow:

Weeks 1-2: Audit Phase

• Export all content with last modified dates
• Identify Tier 1 pages due for refresh
• Check AI visibility for each (Am I Cited)
• Prioritize top 10 for this cycle

Weeks 3-6: Refresh Phase

• 2 articles refreshed per week
• Full substantive updates per checklist
• Updated visuals and examples
• New FAQ questions added
• Editorial review

Weeks 7-8: Tier 2 Audit

• Identify Tier 2 pages due for 6-month refresh
• Prioritize 5-10 for this cycle
• Lighter touch updates

Weeks 9-12: Tier 2 Refresh + Re-promotion

• Complete Tier 2 refreshes
• Re-promote all updated content
• Share on social channels
• Update internal links to point to fresh content
• Newsletter features

Tracking spreadsheet:

The key insight: Build refresh into your content calendar. It’s not ad hoc work - it’s scheduled production.

Technical implementation for freshness signals:

1. Visible date on page:

<time datetime="2025-12-29">Last updated: December 29, 2025</time>

Make it visible to users, not hidden in footer.

2. Schema markup:

{
  "@type": "Article",
  "datePublished": "2024-06-15",
  "dateModified": "2025-12-29"
}

ISO 8601 format required.

3. Sitemap lastmod: Only update when you make substantive changes. Don’t update for every trivial edit. Crawlers learn to ignore the signal if abused.

4. IndexNow notification: Push changed URLs to search engines immediately. Faster discovery of your updates.

5. Page performance: Slow pages = less frequent crawling. Fast pages = discovered updates sooner. Keep Core Web Vitals optimized.

Common technical mistakes:

• datePublished and dateModified are the same (update dateModified)
• Dates in sitemap don’t match page dates
• Modified date visible only in schema, not on page
• Updating lastmod for trivial changes (dilutes signal)

Align all freshness signals for maximum effect.

Different content types need different refresh strategies:

Time-sensitive content (monthly):

• Industry news and trends
• Tool comparisons (features change)
• Pricing information
• Regulatory/compliance content

Evolving topics (quarterly):

• How-to guides for changing software
• Best practices (evolve over time)
• Case studies (add new ones)
• Comparison content

Stable topics (biannual):

• Foundational concepts
• Historical content
• Evergreen tutorials
• Definition pages

What to update by type:

The balance: Authority topics need depth (less frequent, more substantial updates). Fast-changing topics need freshness (more frequent, lighter updates). Match your strategy to topic velocity.

Most people forget the most important part: re-promotion.

Why re-promotion matters: Updated content needs fresh signals:

• New social shares
• New backlinks
• Fresh engagement
• Renewed internal links

Without re-promotion, your update is invisible.

Re-promotion checklist:

• Share on social media (treat as new content)
• Feature in email newsletter
• Update internal links from newer content
• Reach out for updated backlinks
• Submit to relevant communities
• Add to resource roundups

Our re-promotion results: Updated but not promoted: +15% traffic Updated AND promoted: +65% traffic

Practical approach: Schedule re-promotion for every refresh. Put it in your content calendar. Allocate time/budget for promotion.

The compound effect: Fresh content + fresh signals = AI systems recognize genuine relevance. Just updating isn’t enough.

How to monitor if freshness updates are working:

Track before/after:

• AI citation status (Am I Cited)
• Organic traffic (6-week window)
• Keyword rankings for target terms
• Engagement metrics (time on page, bounce)

AI visibility tracking: Test key prompts before and after update. Document which content gets cited. Compare to competitor visibility.

Warning signs you need to update:

• Traffic declining for 3+ months
• Dropped from AI answers
• Competitors appearing instead
• Content age > 6 months without refresh

Our monitoring dashboard:

Guide B is overdue for refresh.

Set alerts:

• Traffic drops > 20% in 30 days
• Disappeared from AI answers
• Competitor visibility increase

Proactive monitoring beats reactive firefighting.

Freshness mistakes that hurt more than help:

Mistake 1: Date inflation without updates Just changing the date fools no one. AI systems detect content hasn’t changed. Can actually hurt credibility.

Mistake 2: Too-frequent tiny updates Adding a sentence every week. Better: Substantial update every 90 days. Quality over quantity of updates.

Mistake 3: Ignoring Tier 1 content Focusing on new content while old pillars age. Your best content needs the most attention. New content won’t save dying pillars.

Mistake 4: No tracking system Can’t manage what you don’t measure. Build a content inventory with dates. Set automated reminders for refresh cycles.

Mistake 5: Updating everything equally Not all content deserves the same investment. Tier your content ruthlessly. Let low-value content age gracefully.

Mistake 6: Forgetting re-promotion Updated content needs distribution. Treat refreshes like new publications. Invest in promotion, not just production.

Mistake 7: Conflicting date signals Page says 2025, schema says 2024, sitemap says 2023. Align ALL freshness signals. Inconsistency confuses AI systems.

This gives me a sustainable refresh strategy. My plan:

Immediate (This Week):

• Build content inventory with all 150 articles
• Add last modified dates and traffic data
• Tier content into 1/2/3 categories
• Identify top 15 Tier 1 pages

Month 1:

• Refresh top 10 Tier 1 pages (2 per week)
• Full substantive updates per checklist
• Re-promote each refreshed piece
• Track AI visibility changes

Ongoing Cadence:

• Tier 1: Every 90 days (rolling schedule)
• Tier 2: Every 6 months
• Tier 3: Annual review

Tracking System:

• Spreadsheet with all URLs and dates
• Calendar reminders for refresh cycles
• Monthly AI visibility checks
• Traffic trend monitoring

Checklist per refresh:

• Add 500+ words new content
• Update all statistics
• Refresh screenshots/examples
• Add new FAQ questions
• Update intro for current context
• Align all date signals
• Re-promote actively

Expected outcomes:

• Month 3: Tier 1 content back in AI answers
• Month 6: 50% visibility improvement
• Year 1: Sustainable refresh rhythm established

Thanks all - this makes content freshness manageable.