Citations Decay. Plan For It.

A citation is not permanent. It has a life span, and most brands never measure it.

Here is what that life looks like. A page gets cited for the first time. Citations climb. They hit a peak. Then, often quietly, they fall. Call the drop-off point the half-life: the day citations fall to half their peak and stay there.

Most content teams ignore this curve completely. They refresh on a fixed schedule instead. Quarterly reviews. Annual audits. The page that needed an update three weeks after publishing sits untouched for months, while a page that is still climbing gets refreshed for no reason at all.

That is backwards. The pages worth fixing this week are not the oldest ones. They are the ones whose citations already peaked and are now falling.

Different formats decay at different speeds. A glossary entry might hold its citations for months. A launch post about a feature might peak fast and fade in weeks. Knowing which is which tells you where to spend your time. Long-half-life formats are worth producing more of. Fast-decaying formats need a different approach, or a much higher publishing rate to keep pace.

Half of what AI engines cite right now is under three months old. That number alone should change how you think about content. Publishing something once is not a strategy. It is the start of a countdown.

The fix is not complicated, but it does require watching the right signal. Track the first-cited date, the rise, the peak, and the half-life for every page that matters to you. Sort by what is falling fastest. Work that list first. Leave the pages still climbing alone.

Waiting for a calendar reminder means you find out a page went stale months after it happened, usually from a customer who got a wrong or outdated answer instead of you.

beket.ai tracks how your citations rise and fall across every model your customers use, so you know exactly which pages to refresh before they go quiet, not after.