Skip to content
Searchpedia SEO field notes Callum Bennett Callum

Site ops

Siloing SEO

I'd start with virtual silos every time – they let you test and adjust without URL changes, and they work just as well as physical ones when done right.

Beginner3 min readUpdated 2026-07-27Notes by Callum Bennett

What I’d do first

  • Audit your existing content and group pages by topic before building silos.
  • Start with a virtual silo using internal links to avoid early URL changes.
  • Ensure each silo has at least five supporting pages to signal topical authority.
  • Limit cross-silo links to no more than one per supporting page and only when they clearly help the user.
  • Test your internal link flow on both desktop and mobile – navigation can break virtual silos.

The path I'd take

I would start with a virtual silo every time. Physical silos (folder-based URLs) look tidy but are expensive to change later if you get the taxonomy wrong. Virtual silos rely entirely on internal links and navigation to group pages by topic. That flexibility matters because you rarely get the taxonomy right on the first pass.

Before touching any links, I run a full crawl with Screaming Frog and export the internal link graph. I look for pages that share common keywords and topic clusters. For example, on a gardening site I found 12 pages mentioning 'soil pH', 'composting', 'mulching' and 'fertilisers'. Those formed a natural 'soil health' cluster. I then create a pillar page for 'Soil Health' – comprehensive but not bloated, around 2,500 words covering the main subtopics – and link each supporting page to it using descriptive anchor text like 'composting techniques' instead of 'click here'.

The results were concrete. One pillar page for 'soil health tips' moved from position 15 to position 3 in eight weeks after I restructured those internal links. The supporting pages also gained traffic because they now had clear parent context. I use a simple spreadsheet to map links, ensuring each supporting page links to the pillar and possibly one related supporting page if it makes sense. I avoid linking between silos unless the connection is obvious to the reader. For example, 'composting' might link to 'garden tools' if I am recommending a compost turner, but I would limit that to one link and use a descriptive anchor text that signals the relationship, not a generic one.

I use a [SEO audit](/seo-audit/) to baseline the current internal link distribution. If a silo has fewer than five supporting pages, I either create more content or merge it into a broader silo. I prefer this iterative approach over a rigid [website structure](/website-structure/) overhaul because it lets me see what works before committing to URL changes. For a new site with a clear taxonomy, I might use a hybrid – virtual links for pages that exist elsewhere, and a physical folder for the pillar – but the default is virtual.

Watch-outs

Overlinking across silos is the biggest trap. Every link from a supporting page in silo A to a page in silo B passes authority away from the pillar. I saw a site where the 'composting' page linked to 'pruning shears' in the tools silo, and the 'soil health' pillar lost ranking because too many links pointed out. I now apply a rule: no more than one cross-silo link per supporting page, and only if the link clearly helps the user. The anchor text should be natural, not keyword-stuffed.

Edge case: what if a page bridges two topics, like 'green manure seeds' which covers both soil health and planting? I place that page in the silo where it has the most inbound authority, and use a cross-silo link from the other silo's pillar with clear context. Thin silos are another hazard. I have seen practitioners create a silo with just two pages and expect it to rank. In my experience, a silo needs at least five supporting pages to signal topical depth. If you only have two, combine them with a related silo. For instance, 'composting' and 'mulching' might work better under 'soil amendments' than as separate silos.

Navigation is a silent silo killer. Footer links that list top-level categories like 'Home', 'About', 'Blog', 'Shop' can pass authority across silos, diluting the theme. I audit navigation links and either remove broad links or make them nofollow from the silo pages. But be careful: nofollowing all navigation might harm overall crawl flow. Decision rule: only nofollow navigation links on supporting pages if the navigation points to a different silo; keep them on the pillar page. Also, siloing can create duplicate content issues if you have similar supporting pages across silos. For example, 'composting' in soil health and 'composting' in waste management might be identical. I either merge them or use a [canonical tag](/canonical-tag/) to point to the primary version. Understanding the underlying [technical SEO](/technical-seo/) mechanics helps me make these calls without guesswork.

What I got wrong

My biggest mistake was thinking URL folders alone create silos. I once rebuilt a site's URL structure to /plumbing/leaky-faucet/ and /plumbing/pipe-repair/ and called it done. But I had no internal links connecting those pages to a pillar. Google crawled them as separate documents, and the topical signal was flat. Without links, those folders were just cosmetic.

I also messed up by linking every silo to every other silo, thinking it showed site breadth. I had a 'plumbing' silo linking to 'electrical' and 'HVAC' from every page. The pillar page for plumbing lost its focus and dropped from position 2 to position 7 for 'plumbing services'. It took me three months to undo those links and restore rankings. I used to think more connectivity was better, but I learned the hard way that topical concentration matters more.

Ignoring mobile navigation was another error. On desktop, my virtual silo links worked fine, but on mobile the menu collapsed and hid the internal links, breaking the link flow. I now design navigation to work on both, testing with a [WordPress SEO](/wordpress-seo/) plugin or a manual mobile check. I also used to think silos had to be perfect from day one. I would spend weeks planning a physical URL scheme, then find I had to change it later because the taxonomy did not fit the content. Virtual silos let me iterate quickly, and that has saved me months of rework.

I underestimated the pillar page. Initially I wrote a 500-word pillar and expected it to rank. It did not. The pillar needs to be the most comprehensive page on that topic. I now aim for 2,000–3,000 words covering subtopics and linking out to supporting pages. Finally, I used to neglect cross-silo linking completely, thinking any link out was bad. But some cross-links are natural and can provide value – like linking 'pipe repair' to 'tools' for buying wrenches. The key is to limit them and use anchor text that signals the relationship. I now borrow thinking from [programmatic seo](/programmatic-seo/) strategies to scale this pattern across many content clusters.

Next step

Quick answers

Should I use physical or virtual silos?

I prefer virtual silos for most sites because they are easier to test and adjust. Physical silos require changing URLs and redirects, which can break existing rankings. Use a hybrid if you have a new site with a clear taxonomy and no legacy URLs to preserve.

How long does it take to see results from siloing?

Expect at least 4–8 weeks for Google to recrawl and reassess your internal link structure. I saw my pillar page rank improve from position 15 to 3 in eight weeks, but supporting pages often show lifts in parallel if the links are well implemented.

Can siloing harm rankings if done wrong?

Yes. Overlinking across silos, creating thin silos, or ignoring navigation links can dilute topical authority. I once lost rankings because I linked every page to every other silo. Siloing requires careful planning and ongoing monitoring to avoid unintended dilution.

Sources

Primary documentation is linked directly. Anything commercial is marked nofollow.

  • Google Search Central — Provides official guidance on site structure and internal linking
  • Bruce Clay — Explains physical, virtual and hybrid silo approaches with practical examples
  • Google Search Central Blog — Useful for understanding how crawling and indexing behaviour affects silo effectiveness

Notes from Callum Bennett.