Content silo

The idea traces back to the early 2000s, when Bruce Clay popularized “siloing” as a way to make a site’s themes legible to search engines. His analogy: a site with mixed, undifferentiated content is like a jar of mixed marbles, and an engine reads subject relevance clearly only when the topics are separated into distinct groups, like a jar of green marbles and a jar of red marbles.

Physical vs virtual silos

There are two ways to build one. A physical silo uses the URL directory structure, so every page about a subtopic lives under a shared path like /power-tools/cordless/. Google’s own starter guide notes that using directories to group similar topics helps it learn how often a section’s pages change, so it can crawl each section at the right cadence. A virtual silo does the same job through internal links: contextually connecting related pages so the theme reads clearly even when the URLs are not neatly nested.

Classic siloing told you to keep silos isolated and avoid linking between them. Modern practice is looser. Rigid isolation creates orphan pages and blocks useful cross-links, and search engines now reward natural, helpful linking over artificial structure. That is why most teams run topic clusters (a pillar page plus interlinked supporting pages) rather than sealed silos. The goal is unchanged: concentrate relevance so one strong theme reads as depth, not a scatter of one-off posts.

A silo is a content-architecture decision, so it lives inside broader on-page work: URL structure, headings, internal anchor text, and how each page maps to search intent.