Skip to content
DirectoryFast
All posts

SEO Site Structure for Directories: URLs, Categories and Click Depth

The skeleton of a directory: URL hierarchy, categories and sub-categories, click depth and pagination. One URL per listing, collection pages that earn their place, and Google's pagination rules applied to directories.

Every directory has a structure, whether you designed it or not. The URLs you picked on day one, the way categories nest, how many clicks it takes to reach a listing: Google reads all of it, and so do your visitors.

This guide covers the skeleton of a directory: URL hierarchy, categories and sub-categories, click depth and pagination. It is not about which pages link to which, which is a separate topic covered in Internal Linking for Directories. Get the skeleton right first, because changing it later means redirects, and redirects on thousands of pages are expensive.

The three layers every directory needs

Almost every directory, whatever the niche, fits into three layers:

  1. Hub pages. The home page and top-level categories. "Coworking spaces", "Design agencies", "MCP servers".
  2. Collection pages. Sub-categories, locations, or category plus attribute. "Coworking in Lisbon", "Design agencies for SaaS", "MCP servers for databases".
  3. Entity pages. One listing each. One coworking space, one agency, one server.

Each layer answers a different kind of search. Hubs answer broad queries, collections answer the specific "X in Y" or "X for Y" queries where directories tend to win, and entity pages answer searches for a named thing.

Problems start when the layers blur: a listing that also acts as a category, a location page that duplicates a sub-category, or five ways to reach the same collection.

URL structure: pick one pattern and stick to it

Google's URL structure guidelines recommend simple, descriptive URLs with readable words rather than long ID numbers, and hyphens rather than underscores to separate words. For a directory, that usually leads to one of two patterns.

Nested:

/coworking/
/coworking/lisbon/
/coworking/lisbon/example-space/

Flat entities:

/coworking/
/coworking/lisbon/
/listing/example-space/

Nested URLs read well and show the hierarchy in the address itself. Their weakness shows when a listing belongs to two collections (a coworking space that is also a café, say). You then need to pick one canonical path, or you end up with two URLs for one listing.

Flat entity URLs avoid that problem. The listing has one address forever, no matter how many categories it appears in. The hierarchy lives in the breadcrumbs and links, not in the path.

Neither is "the SEO-correct one". What matters:

  • One URL per listing. If a listing can be reached at two paths, one of them needs a canonical pointing to the other, or a redirect.
  • Stable slugs. Don't put things that change (ratings, prices, years) in the slug. A changed slug means a redirect.
  • No content behind fragments. Google's guidelines say not to use fragments (the part after #) to change the content of a page, because Google Search generally doesn't support them. Filters and tabs that only change the fragment won't be seen as separate pages.

Categories: depth is earned, not designed

The temptation on a new directory is to build the full taxonomy up front: 12 categories, 60 sub-categories, 400 locations. Most of those pages then go live with one or two listings each.

A thin collection page answers its query worse than the single listing on it. It is also the kind of page that ends up in Search Console under "Crawled - currently not indexed".

A better rule: a collection page exists when it has enough listings to be useful. Pick a threshold that fits your niche. Below it, the listings stay visible on the parent category, and the sub-category page is either not created yet or kept out of the index.

Some practical habits:

  • Name categories the way people search, not the way your database labels them.
  • Avoid overlapping categories. "Remote jobs" and "Work from home jobs" on the same site split the same intent across two pages.
  • Treat locations and attributes as collections only when the data supports them. "Coworking in Lisbon with 24/7 access" is a good page if you actually record 24/7 access for every listing.

Click depth: keep important pages close to the home page

Click depth is the number of clicks from the home page to a given page. There is no official Google number to hit, but the logic is simple: pages buried deep in the structure get fewer internal links, and look less important to visitors and to crawlers.

For most directories, a sensible target is:

  • Home to category: 1 click
  • Category to collection: 1 click
  • Collection to listing: 1 click

So every listing sits about three clicks from the home page. If you find listings at six or seven clicks, the cause is almost always pagination (see below) or a missing collection layer.

How to check it: crawl your own site with any SEO crawler, sort pages by depth, and look at which templates sit at the bottom. Then compare that list with the pages Search Console reports as not indexed. The overlap is often large.

Pagination: every page needs its own URL

Large categories need pagination, and pagination is where directory structures leak the most. Google's guidance on pagination is precise:

  • Link pages sequentially with normal <a href> links from each page to the next.
  • Give each page a unique URL, for example with a ?page=n parameter. Google treats pages in a sequence as separate pages.
  • Don't use the first page as the canonical for the whole sequence. Each page gets its own canonical URL.
  • Don't use fragments for page numbers. Googlebot may not follow a link that only differs by the text after #.
  • Google no longer uses rel="next" and rel="prev", although other search engines may.

Two directory-specific additions:

  • Sort pages so that deep pages still get links. If listing 340 of a category is only reachable through page 17, give it another route: a sub-category, a location page, or "related listings" from its neighbours.
  • Don't index filtered or re-sorted versions. Google's documentation recommends avoiding indexing URLs with filters or alternative sort orders. Your canonical category page is the one to rank.

Breadcrumbs make the structure visible

Breadcrumbs show both visitors and crawlers where a page sits: Home › Coworking › Lisbon › Example Space. They also add one link per level, on every page, without any extra work.

Mark them up with BreadcrumbList structured data. The markup does not change rankings by itself, but it helps Google understand the hierarchy and can appear in search results. The schema markup guide for directories covers the code.

A short structure audit

If your directory already exists, run this once:

  1. List every page template (home, category, collection, listing, tag, filter, blog).
  2. For each template, count URLs and count how many are indexed in Search Console.
  3. Crawl the site and note the average click depth per template.
  4. Find listings reachable through more than one URL.
  5. Find collection pages under your listing threshold.

Most of what you find will be fixed by merging thin collections, canonicalising duplicate paths, and adding one collection layer where categories are too large.

FAQ

Should directory URLs include the category?

They can. Nested URLs are readable, but they force you to choose one category per listing. If listings often belong to several categories, flat entity URLs with breadcrumbs are easier to keep clean.

How many clicks deep is too deep?

Google doesn't publish a number. As a rule of thumb, keep important listings within about three clicks of the home page, and investigate any template that sits much deeper.

Do I need a page for every location?

No. Create a location page when it has enough listings to be useful. Until then, show those listings on the parent page.

Should paginated pages be noindex?

Google doesn't ask for that. Its guidance is to give each page in the sequence its own URL and canonical and to link them in order. Filtered and re-sorted versions are the ones to keep out of the index.

Is changing my URL structure risky?

It is work, not a disaster, if you redirect every old URL to its new equivalent. See Migrating a Directory to a New Domain Without Losing Rankings for the redirect checklist, which applies to URL changes on the same domain too.

Where to go next

Structure is the part of directory SEO you set once and live with. Decide on one URL per listing, earn your collection pages with real listings, and keep every listing a few clicks from the home page. Then move on to internal linking, which builds on this skeleton.

Run a directory? DirectoryFast is building an indexing check for directory sites: connect Search Console and your sitemap, and see which pages to fix, by template. See what DirectoryFast is building →

Related reading

Sources