Your Old Content Is Losing Ground Faster Than You Think
Content decay existed before, but the timeline has shrunk with the rise of AI Search. Airops data shows that pages not updated for 3 months are now 3× more likely to lose visibility compared to content that has been recently refreshed. Rather than a slow fade, it’s more of a hard drop.
Not so subtle shift: Many AI search engine, including Google AI Mode and AI Overviews which help users search for timely information, view freshness as a key ranking signal related to a searcher’s intent. And, this pressure isn’t just applicable to news-related queries. Rather, it’s for all types of queries. This means your pages are not just competing for rankings. They’re also competing for citations. And when it comes to citations, AI engines generally choose fresher pages.
” Honestly, this is more a mindset thing than anything else – it’s realising that the field is changing drastically and what was true before is not true anymore”
— Abhishek Iyer, founder of ACME.BOT
The gap’s widening fast. out-of-date pages are falling behind—and it seems there’s a strong bias everywhere, on every query type, and every platform.
Why AI Search Punishes Stale Content More Than Traditional Google Ever Did
When ranking content, Google would only apply “query deserves freshness” mainly for breaking news or time-sensitive topics. AI search, on other hand, treats freshness as a universal signal across nearly every query type.
Here’s an analogy: traditional Google is like a librarian. It indexes a lot of information and has the ability to pull older books when needed. AI search is like a news editor — it’s largely looking for current day information, not data from a day or year ago.
Analysis of 16.975 million cited URLs across 7 AI platforms by Ahrefs shows that AI-cited bursts in content freshness, with content being 25.7% fresher on average than organic SERP results. That’s a gap that compounds fast.
Ninety days out from publishing, our Pages with AI Overview see almost a 40% drop in their AI Overviews citation rate, despite the underlying facts remaining unchanged. Even though the Overview itself is still accurate, their citation rate drops.
Content that’s under 30 days old gets roughly 3.2x more AI citations than older pages. That’s the freshness multiplier you’re competing against right now. This isn’t about updating just this content for users anymore—it’s about staying in front of the routing systems that send traffic to you.
What Data Tells Us About AI Mode and Old Content
The citation data is not ambiguous: In Meltwater’s data examining 9 million citations, nearly half of the successful AI citations were published within the last 3 months, and only 12% were more than one year old. It’s not a trend toward more recent content, but a wholesale rotation.
What this means in practice: Each quarter, AI engines such as Google AI Mode recycle their citation pool. This means they recycle or scan through the same list of potential citations. It’s not that the older ones aren’t valid, but because fresher alternatives exist.
A study by Seer Interactive found that 75% of pages cited in LLMs were updated in the last year. Just when a page was updated can be as important, or in some cases, more important, than when it was first published.
Content with the freshness factor older than 14 days registers a 23% reduction in AI citation frequency.
According to Search Engine Land, less than 1 in every 3 Google searches send people to the open web by clicking on a link, and AI Overviews alone reduce CTR on organic results at the top by an estimated 58%. So, while citations are drying out, so is the traffic that accompanied them.
This isn’t a temporary blip.
Content decay with AI search is no longer a fluke, but a structural shift—the disappearance of old content from Google’s results is a structural shift, not an algorithm hiccup. Which raises an entirely different question: not whether to refresh, but how often.
How to Fight Content Decay and Stay Visible in AI Search
AI search buries old content even faster than traditional Google ever did. Here’s what works.
1. Audit declining pages first
Generate a report of your 50 highest-performing pages and scan for pages losing organic traffic or AI citations. Hit the pages that could still be “saved” first.
2. Update the
Freshness AI engines primarily treat
3. Go quarterly, not annual
Think quarterly, not annual. Quarterly refreshes have been shown to provide 42% better results than annual ones, according to Animalz. A single refresh even brought a 55% weekly traffic lift. That’s the difference between remaining relevant and fading into obscurity.
4. Write self-contained sections with strong headers
For AEO, LLMs pull individual chunks — not entire pages. Make sure that each chunk includes a strong header that answers fan-out questions individually.
“In my opinion, the single most important metric for AI search is the visibility of long-tail / fan out queries. No longer is it enough to simply rank for the main query — you also need to rank for fan out queries.”
— Abhishek Iyer, founder of ACME.BOT
5. Treat content as a living asset
Brands that continuously publish and refresh content are more likely to hold stronger AI citations over time. One quarterly refresh can have compounding, real-world results.
Flourishing in the Long Game with Freshness, Authority, and the Right Mindset
There is no skipping the basics. Organic search powers it all — if a search engine doesn’t rank the page, it won’t ever be cited by AI. Because LLMs are non-deterministic and they’ll return different answers to each user, it’s better to focus on topical authority and fan-out queries instead of studying the number of AI citations you’re getting for any one query.
“IMHO, tracking AI citations directly at this point is meaningless. Tech and LLMs are evolving quicker than you or I can keep up with. Every user sees a different result. Also, LLMs are non deterministic.”
— Abhishek Iyer, founder of ACME.BOT
The work we do at ACME.BOT to track and analyze patterns of content decay shows us that consistent refreshes that align with principles for answer engine optimization — as opposed to one-off rewrites — maintain LLM search visibility. At one point, one refresh generated a 55% weekly traffic lift. That’s the payoff for treating content as a living asset.
