Collection Page SEO Checker

Why the category page is the head-term asset

A product page ranks for one product's name. A category page ranks for the phrase a shopper types before they know which product they want: running shoes, standing desks, matte lipstick. Those head terms carry the volume, and the category page is the only URL on the store built to answer them.

Google's ecommerce site structure guidance puts the category page between the menu and the catalogue: link from menus to category pages, from those to sub-categories, and from sub-categories to every product. It also warns that if category pages do not link directly to all products, Googlebot might not find them by crawling alone.

That gives a category page three jobs at once, and the checks on this page follow them:

• Rank for the head term. The title tag, H1, meta description and the category's own copy are the on-page signals that say what the page is about.
• Pass discovery to products. Product links are internal links Googlebot follows, and pagination decides whether products past the first batch have a path at all.
• Stay the one URL for the job. The canonical, the robots directives and the filter URLs decide whether the category itself is indexed, or one of its hundred sorted and filtered copies is.

Google's ecommerce URL structure guidance adds the part stores get wrong most: each paginated page needs a unique URL, parameters should be key=value rather than bare values, and internal links should never carry session IDs or other temporary values. The category page SEO guide covers the copy and layout side; this tool reads the technical side.

What each check means and the threshold it uses

Every check reports the value the page served next to its verdict, so you can see the 193-character description rather than be told it is long. The thresholds are below, and the reference table after them says what a pass means for each row.

Title and meta description length

A title passes between 10 and 60 characters and a description between 50 and 160, the same thresholds as the meta tag checker so the two tools never disagree about one page. Over the limit is a warning because Google trims it. Missing is a fail because Google then writes one from a product card.

Self-referencing canonical

The canonical tag on a category page should name itself. A filtered or sorted URL canonicalising to the category also passes. Pointing at another URL is a warning, two tags is a fail, and a later page pointing at page 1 is a fail: Google's pagination guidance says not to use the first page as the canonical.

Pagination

Google's pagination documentation asks for a unique URL per page and states that Google no longer uses rel=next and rel=prev. So the checker scores plain links to page 2 and records the rel attributes as information. A page with rel=next but no anchor to page 2 is a warning: products past the first batch have no crawlable path.

The category description heuristic

Category copy is text that belongs to the category rather than to a product card, and no HTML element marks that boundary. The tool reads it in two passes and tells you which one it used:

• Known containers first. Shopify's collection-description and .rte inside the collection header, WooCommerce's term-description and products-header description, and the category description blocks BigCommerce and Magento themes print.
• Then paragraphs outside cards. Text in p elements inside main, skipping anything inside a product card, list item, article, navigation, site header, footer, aside or form.
• 50 words passes. Fewer is a warning, not a fail, because a description the heuristic cannot see is possible and the report says so.

How to read a pass, warn and fail report

The banner counts the scored rows. Informational rows are listed but never counted, so a page can show two fails and still have thirteen rows. Read the report top to bottom, because the rows are ordered from the signals that decide whether the page ranks at all to the ones that decide how well.

• Fail means the page cannot do its job as served: no title, no description, no H1, a noindex, conflicting canonicals, page 2 pointing at page 1, or a grid with no product links.
• Warn means it works but leaves something on the table: a long title, a missing canonical, thin copy, no breadcrumb markup, or product images without alt text.
• Pass means the observed value meets the threshold in the table above. It does not mean the value is good copy, which is why it is shown.
• Info means the row is a fact about the page, such as the filter URL count or the platform, that needs your judgement rather than a rule.

Two results deserve a second look. A description at 0 words on a page that has copy means the copy sits in a container the heuristic does not know, or JavaScript renders it. A product-link fail on a page where you can see products means the same about the grid, which the technical SEO for ecommerce guide covers.

What to do after the report

Most rows are template problems, which is good news: the canonical, breadcrumb markup, pagination links and alt handling are set once in the collection template and fixed for every category. The title, description and copy are not, and those are where the ranking work is.

1. 01

   Paste the first page of the category

   Use the URL shoppers land on from the menu, without ?page=, ?sort_by= or a filter parameter. Page 1 is the URL that should rank for the head term, and the tool reads it as served, so a filtered or paginated URL produces a report about that variant instead.

2. 02

   Read the platform and the fetch line

   The banner names the platform it recognised and the URL it landed on after redirects. If the platform is unknown, the product-link and description checks fall back to generic patterns, and a redirect to a different host means you checked a URL the store no longer serves.

3. 03

   Clear the fails first

   A fail is a missing title or description, no H1, noindex, two canonical tags, page 2 canonicalising to page 1, or a grid with no crawlable product links. Each one stops the page ranking or stops products being found, and each is a template fix that repairs every category at once.

4. 04

   Then work the warnings in the order they appear

   Title and description length, a missing canonical, no breadcrumb markup, thin category copy and product images without alt text are all things a competitor's category page probably has right. The observed value column tells you what to change without opening the page source.

5. 05

   Read the informational rows for crawl budget

   Filter and sort links and the schema row are recorded, not scored. Dozens of crawlable filter URLs are fine when they canonicalise to the category; thousands with no canonical or robots rule are a crawl budget problem the checker cannot see from one page but this count hints at.

6. 06

   Rerun, then check the sibling categories

   Most fixes live in the collection template, so one change repairs every category and a rerun confirms it. The description and the title do not: those are per category. List your head terms, map one to each category page, and check each URL once the copy is in.

To confirm one fix, the canonical tag checker validates a single URL's tag in isolation, and the breadcrumb schema generator writes the BreadcrumbList block a category page is missing. The ecommerce SEO checklist puts these checks in order with the rest of the store.

What this tool checks

Frequently asked questions