> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getcited.in/llms.txt
> Use this file to discover all available pages before exploring further.

# Recency Bias

> The tendency of LLMs — especially retrieval-augmented ones — to favor recently published or updated content.

**Recency bias** is the tendency of [LLMs](/glossary/llm), particularly those using [retrieval-augmented generation](/glossary/rag), to favor recently published or recently updated content over older pages on the same topic. A page updated this month is more likely to be retrieved and cited than an equivalent page last updated two years ago.

## Why it matters

Recency bias creates both an opportunity and a risk. The opportunity: brands that publish or refresh content regularly gain a systematic advantage over competitors with stale pages. The risk: brands that stop updating lose AI visibility gradually as their content ages out of retrieval preference. [Perplexity](/concepts/platforms/how-perplexity-ranks-sources) shows the strongest recency bias of the tracked platforms; [Claude](/concepts/platforms/how-claude-search-works) shows the least because it relies primarily on training data.

## How it applies in practice

Cited tracks [content freshness](/glossary/content-freshness) as a diagnostic factor in [mention rate](/glossary/mention-rate) analysis. When a brand's visibility drops without a clear content or competitive cause, stale content is one of the first hypotheses. The recommended response varies by category: fast-changing verticals (tech, fintech) benefit from monthly content refreshes, while stable categories may sustain visibility with quarterly updates.

## Related concepts

* [Content freshness](/glossary/content-freshness) — the metric recency bias acts on
* [How Perplexity ranks sources](/concepts/platforms/how-perplexity-ranks-sources) — the platform with the strongest recency bias
* [Refresh cadence](/methodology/refresh-cadence) — how often Cited's own data pipeline runs