Content Decay: Why Your Blog Posts Disappear From AI Search Every 13 Weeks

Content decay is the gradual loss of AI citation frequency that happens when content is not updated. Roughly half of the content AI search engines cite was published or updated in the last 13 weeks (Amsive, 2026). So a page left untouched for longer than 13 weeks fades out of AI answers, even while it still ranks on Google. The fix is a 13-week refresh cycle on your highest-value pages.

A guide can rank on page one of Google for three years and still be cited by Perplexity exactly zero times. The ranking says the page is good. The silence says it is stale. Those are two different judgements, made by two different systems, and in 2026 the second one decides whether buyers researching on AI ever see you.

This is the gap a content library falls into when it is left alone. The pages were written well, ranked early, and got indexed by AI engines in their first weeks. Then nobody touched them. And with every week that passed, their odds of being cited dropped.

A content calendar without a refresh schedule is half a GEO strategy.

What is content decay?

Content decay is the gradual decline in a page's effectiveness over time. In classic SEO it meant ranking slippage as fresher competitor pages displaced yours. In AI search it is faster and harsher: citation frequency falls away once a page passes roughly 13 weeks without an update.

The mechanism is simple. AI engines treat freshness as a proxy for accuracy. A statistic from 2023, a market description from 2022, a list of "current" best practices written before AI search was mainstream: each one reads as an accuracy risk to an engine. Given two comparable sources, the engine cites the more recently updated one with more confidence.

That creates a structural problem for any business that built a content library without a refresh schedule. Its best pages, well written and ranked and indexed, are quietly losing AI citation frequency every week they sit still.

Where does the 13-week number come from?

The threshold comes from how AI retrieval systems weight recency. Amsive's 2026 analysis found that roughly 50% of all content cited by AI search engines was published or updated in the last 13 weeks. Independent 2026 datasets point the same way: AI-cited URLs run materially fresher than the pages ranking in Google's organic top 10, and content under three months old is cited at a much higher rate than older pages.

Thirteen weeks is not a cliff where citations stop. It is the point where the probability of being cited drops noticeably and keeps declining. A page that was earning regular AI citations six months ago can be earning far fewer today for one reason only: it has not been refreshed.

The practical implication is concrete. Every page you care about being cited needs a refresh date set 13 weeks from its last update, and a process to actually update it when that date arrives.

What does content decay look like in practice?

Content decay is invisible unless you are measuring AI citation frequency, and almost no business currently is.

Picture a professional-services firm that publishes a Cyprus company-formation guide in Q4 2024. Through early 2025, AI engines cite it for relevant queries and AI referral traffic is measurable. By Q3 2025, with no change to the page, citations start sliding. By Q1 2026 the guide barely appears in AI answers, while still ranking strongly on Google.

The SEO dashboard shows a healthy post. Rankings are flat. Organic traffic is flat. Nobody sees the AI decline, because nobody is tracking it. The content is decaying in silence.

That split between Google performance and AI performance is now one of the most common visibility problems for businesses whose libraries were built in 2022 to 2024. And it is a live risk locally. In Axenor's June 2026 audit of 10 Cyprus corporate-services firms, seven had no proper structured data, none had a Wikidata or Wikipedia entity, and not one had FAQ markup. Those are exactly the signals an engine leans on once a page's freshness fades. When the freshness signal weakens and the structural signals were never there, the page has nothing left to be cited on.

Which content decays fastest?

Not all content decays at the same rate. Knowing which types rot first lets you prioritise the refresh schedule.

1. Statistical content. Anything citing specific figures, such as "AI search usage grew to X%" or "four in ten travellers now plan trips with AI", has a built-in expiry date. A 2024 statistic quoted in 2026 lowers an engine's confidence in the whole page.

2. "Current state" guides. Anything framed as "the current state of X" or "what you need to know about X in 2025" decays the moment that date reference goes stale. Engines read dates and weight them against freshness.

3. Best-practices content. Best practices change. An engine asked for current GEO best practices will not cite a 2023 article with conviction, because the field, and the model's own understanding, has moved on.

4. Tool and platform content. Anything referencing specific tools or platform features (ChatGPT, Perplexity, Google AI Overviews) decays fastest, because the platforms themselves change every quarter.

How do you refresh a piece of content?

A refresh is not a rewrite. It is a targeted update that restores freshness signals and strengthens extractability. Axenor's standard refresh for a roughly 1,500-word article runs about 60 to 90 minutes.

Step 1. Update every statistic. Find each figure. Check whether newer data exists. Update the number and the year. If the original source has nothing current, find an equivalent recent source and cite it. "A 2023 study found 45% of buyers..." becomes "A 2026 report found [current figure] of buyers...".

Step 2. Rewrite or strengthen the Direct Answer Block. If the page has a Direct Answer Block (30 to 80 words at the top), check it still represents the best current answer to the core question. Rewrite it with current language, a current data point, and the year referenced. If the page has no Direct Answer Block, add one. This is routinely the single highest-impact change in a refresh.

Step 3. Update the FAQ section. Add one new FAQ that answers a question which has emerged since the page first published. That signals the section reflects current understanding, not 2023's questions. Review the existing answers and fix anything outdated.

Step 4. Add a visible "Last updated" date. Put "Last updated: [Month Year]" at the top. It is a direct freshness signal read by engines and humans alike.

Step 5. Update the published or modified date in your CMS. Set it to the refresh date so AI crawlers reassess the page's freshness. Move the date only when you have actually changed the content (see the FAQ on cosmetic date edits below).

How do you build a content refresh calendar?

A refresh calendar is a simple spreadsheet tracking: article title and URL, original publish date, last refresh date, next scheduled refresh date (13 weeks out), and a priority level keyed to AI citation value.

To decide what to refresh first:

1. List your top 10 pages by organic traffic.
2. Run each through an AI visibility test: ask ChatGPT, Gemini and Perplexity the core question the page answers, and note whether you are cited.
3. Pages that rank well on Google but are not cited by AI are priority 1.
4. Pages already cited by AI are priority 2: hold them on the 13-week cycle.
5. Low-traffic pages with no AI citation are lower priority.

For a library of 30 articles on a 13-week cycle, that works out to refreshing roughly two to three articles a month, or two to four hours of maintenance. The compounding citation benefit is worth well more than the time.

Why is refreshing better than writing new content?

A refresh has a structural advantage new content cannot match: it inherits existing authority.

A page indexed for two or more years has accumulated domain authority, any earned links, and a history of user signals Google uses to calibrate trust. Refreshing it to be AI-citable again draws on all of that immediately. New content starts at zero. It has to be indexed, accumulate authority, and establish itself as the relevant resource, which takes three to six months for a typical page.

So a refreshed high-authority page can reappear in AI answers within roughly four to six weeks of being updated. Independent 2026 reporting puts the reappearance window at two to four weeks for pages crawled weekly, stretching to six to eight weeks for pages on a slower crawl. A brand-new page targeting the same query may take six months to reach the same citation frequency.

That is why the GEO play for a business with an existing library is refresh-first, create-second. Create new content to fill genuine gaps. Refresh existing content to defend and extend the AI visibility you already earned.

Frequently asked questions

How do I know if my content is being cited by AI engines? Run test prompts monthly in ChatGPT, Perplexity and Gemini for the core questions your pages answer, and note whether your site is cited. Asking "Where can I learn more about [topic]?" tends to surface the exact source an engine is drawing from. Track the results in a simple spreadsheet.

Does every page need a 13-week refresh cycle? No. Stable topics such as fundamental concepts and historical information can run on a 26-week cycle. Fast-moving topics such as AI search, platform features and market statistics should be on 13 weeks or tighter.

Does refreshing content hurt SEO rankings? No. Updating data, the date, and structure consistently improves SEO rather than harming it. The only real risk is changing the content angle or cutting sections that drive rankings, which a targeted refresh (data and Direct Answer Block) does not do.

What is a Direct Answer Block and why add one to every post? A Direct Answer Block is a 30 to 80 word passage at the top of a page that completely answers the core question and stands alone as an extract. It is the passage AI engines are most likely to lift as the answer. Adding one to a page that lacks it is routinely the highest-impact change in a refresh.

Does updating the date without updating the content help? No. Engines judge freshness on more than the published date: whether statistics are current, whether references are recent, whether the page reflects current understanding. A date change with no content change reads as cosmetic to crawlers and delivers little freshness benefit.

How fast does a refreshed page come back into AI answers? For pages crawled weekly, a substantive update can begin reappearing in AI citations within two to four weeks; pages on a slower crawl can take six to eight. Plan on roughly four to six weeks before a refresh shows up reliably across engines.