83%
What it means
A page that ranked well last year can slide this year. Prices change, new products launch, and competitors publish newer pages. That slow slide is content decay.
Why it's faster in AI search
AI engines strongly prefer fresh pages for buying questions. In our study, 83% of the pages Gemini cited had "2026" in the title or URL, and ChatGPT added the year to every web search it ran.
So a "best X in 2025" page can drop out of AI answers on the first day of the new year, even if nothing else changed.
How to spot it
- Falling traffic on pages that used to do well.
- Fewer AI citations for pages that used to be cited.
- Old facts: prices, screenshots, product names.
How to fix it
Update the facts, add what's new, remove what's gone, then update the date and title. Do the pages that matter most first.
Example
A software blog's "Best help desk tools" page was cited in many AI answers in 2025. In January it stops being cited.
The team checks every price, adds two new tools, removes a dead one, and puts 2026 in the title. Within weeks, it's cited again.
Frequently asked questions
Does every page decay?
No. Pages about stable topics last for years. Pages about products, prices and "best" picks decay fast.
Is changing the date enough?
No. Readers and AI notice stale prices. Update the content, then the date.
How often should I check for decay?
Every few months for important pages, and at the start of each year for anything with a year in the title.