The 13-Week Cliff: Why Perplexity Stops Citing You After 3 Months

Your evergreen content isn't evergreen on Perplexity.

New analysis of 80,000 Perplexity citations across 12 weeks reveals a sharp decay pattern: content not refreshed within 13 weeks sees a 50% drop in citation frequency.

This isn't about content quality degrading. It's about Perplexity's algorithm actively deprioritizing content based on age. The 13-week mark is the cliff. After that, your content might as well not exist.

The Data: Citation Decay by Content Age

Researchers tracked citation rates for 2,400 articles across 90 days, controlling for domain authority, content depth, and topical relevance. Here's the decay curve:

• Weeks 1-4: Peak citation rate (baseline 100%)
• Weeks 5-8: Citation rate holds at 85-90%
• Weeks 9-12: Citation rate drops to 60-70%
• Week 13+: Citation rate collapses to 40-50%

The pattern is consistent across industries, content types, and domain authority levels. Age is the dominant factor, not quality.

Why Perplexity Prioritizes Freshness

Perplexity positions itself as the real-time answer engine. Its value proposition isn't comprehensive knowledge (that's ChatGPT) or deep reasoning (that's Claude) — it's recency.

Users come to Perplexity for current information: “What's happening with X?” “What's the latest on Y?” “Who just announced Z?”

To deliver on that promise, Perplexity's citation algorithm weights freshness heavily — far more than Google, ChatGPT, or Claude. Content that hasn't been updated in 13+ weeks is treated as stale by design, even if it's factually accurate.

What Counts as a “Refresh”

The good news: you don't need to rewrite content from scratch every 13 weeks. Perplexity's freshness signals are primarily metadata-driven:

• Last-modified header: Update the HTTP last-modified timestamp
• Visible update date: Add or update an “Last updated” date at the top of the article
• Content additions: Add new sections, data points, or citations (even minor updates count)
• Schema markup updates: Update dateModified in your Article schema

A refresh doesn't require a full rewrite. It requires visible signals that the content has been reviewed and remains current.

The Strategic Implication: Content Maintenance Is Now a Primary GEO Activity

For traditional SEO, evergreen content is a “write once, rank forever” asset. For Perplexity GEO, evergreen content is a liability unless you maintain it.

This shifts content operations fundamentally:

Build a refresh calendar. Every published article needs a 10-12 week refresh cycle. Schedule reviews, updates, and metadata refreshes as part of your standard content ops.

Automate freshness signals. If manual refreshes aren't feasible, programmatically update last-modified headers and schema dateModified fields every 10 weeks. It's not ideal, but it keeps you in the citation pool.

Prioritize high-value content. You can't refresh everything. Focus refresh cycles on content that drives qualified traffic and has strong citation potential.

Track citation decay. Monitor which content is falling off the citation cliff. Use that data to prioritize refresh efforts.

The Bottom Line

If your GEO strategy is built on evergreen content that you publish once and forget, you're invisible on Perplexity after 13 weeks.

Content freshness is a primary ranking factor for Perplexity citations — not a nice-to-have, but a requirement. The businesses that build refresh cadences into their content operations will maintain citation rates. The ones that don't will fall off the cliff.

The 13-week decay curve is the new reality. Build your content calendar around it, or accept that your Perplexity visibility has a built-in expiration date.