Content Groups in GSC: How to Analyse Sections, Not Just Pages
Your site's average hides which sections earn the growth. How to build content groups over GSC data with page filters and regex, and read them like a portfolio.
Search Console has no content-groups feature, but its page filters get you the same analysis: define groups of URLs (by section, template, topic cluster, or publish era), read each group as its own trend, and stop letting the site-wide average blend a growing section with a dying one. The method:
- CUT THE SITE INTO GROUPS: By URL section for most sites (/blog/, /docs/, /product/), by regex for templates and clusters. Groups should follow the questions you ask, not the folder structure.
- FILTER PAGES, NOT QUERIES: In GSC Performance, the Page filter with 'URLs containing' or a custom regex is the whole mechanism. Save the patterns; retyping them is how the habit dies.
- READ GROUPS LIKE A PORTFOLIO: Clicks per group over time, clicks per page within the group, and each group against its own history. One growing section can hide two flat ones in the blend.
- ACT AT GROUP LEVEL: A section declining together has a shared cause: a template, an intent shift, a format losing to competitors. Fix the cause once, not page by page.
Twenty minutes to define the groups once, then the section-level read joins your monthly routine. It is the analysis layer between one page and the whole site.
CrawlRaven maintains this grouping as topic clusters: your pages grouped by topic over daily Search Console data, with cannibalization flags and the crawl joined on. This guide is the manual version using nothing but GSC's own filters, which is also how you will sanity-check any tool's groups. Try CrawlRaven free: 1 site, no credit card →
Ask a site-wide traffic chart "how is the content doing?" and it answers with an average. Averages are where insight goes to die: a blog section growing 40% and a template section quietly collapsing produce, together, a flat line and a shrug.
Content groups are the fix, and Search Console will not build them for you: there is no grouping feature. But its page filters do the whole job manually. Here is the version I run: define the groups once, save the filters, and read the site like a portfolio.
Why the site-wide average lies
Three structural reasons the blended number misleads, no matter how honest everyone is:
- Wins and losses cancel. A growing section and a declining one sum to "stable", which is the one word that triggers no action.
- Big sections drown small ones. A 2,000-page archive's noise swamps a 30-page money section's signal in any combined view.
- Causes are sectional. Templates, topics and playbooks fail as units. Per-page analysis discovers the same template bug forty separate times.
Step 1: Cut the site into groups worth reading
Groups should follow the questions you ask, not the CMS's folder habits. Four cuts cover most sites, and three to six groups is the working range:
Content groups, and the question each one answers
In prose: cut by section to learn which part of the site earns growth, by template to judge template changes, by topic cluster to judge topic investments, and by publish era to compare the new playbook's output against the old one's. Past ten groups the portfolio turns back into pages.
Step 2: Build each group as a saved page filter
The mechanism is one filter:
- In the Performance report, choose + New → Page.
- For folder-shaped groups, URLs containing with the path fragment:
/blog/,/docs/. - For everything else, Custom (regex) on the page dimension.
- Save the patterns in a shared doc. GSC forgets your filters between sessions; the doc is what makes the groups an institution instead of one analyst's habit.
Regex patterns for groups that are not folders
- A template: category pages as
/category/[^/]+$, detail pages as the deeper path shape. The page dimension takes the same RE2 syntax as the query filter. - A topic cluster: list the slugs:
striking-distance|cannibalization|gsc-data. Longer to write, but it groups by meaning rather than by URL accident. - Test before trusting. Our free GSC regex generator validates the pattern against sample URLs, which catches the unescaped dot before it silently over-matches.
Step 3: Read the groups like a portfolio
With the filters saved, the monthly read is three numbers per group, each against the group's own history:
- Clicks and impressions per group, the headline trend.
- Clicks per page within the group, which separates a section that grew from a section that merely got bigger.
- Each group's top queries, because a group's character changes before its totals do.
Four patterns a per-page view never shows
Publishing forty new pages grows a group's total while its clicks per page falls, which is a content program producing volume, not value. The division takes one spreadsheet column and regularly reverses the story the total tells.
Step 4: Act at the group level
The payoff for grouping is that fixes get cheaper, because causes are shared:
- A group declining together points at its template or its format. Diagnose one representative page deeply, fix the template once, and the whole group inherits it.
- A group with volume and no value starts the consolidation and pruning conversation with evidence instead of vibes.
- The winning group is the playbook. Study what its pages share, then point next quarter's content at it, and push its almost-ranking pages first via the striking distance loop.
Tips that keep groups useful
- Keep groups stable across quarters. Redefining them every review resets every trend. Change definitions deliberately and note the date in your annotations.
- Prefer clean partitions for reporting. Overlapping groups are fine for exploration, but a page counted twice makes the portfolio arithmetic unexplainable in a meeting.
- Give the money section its own group however small it is. The 30 pages that convert deserve a line the 2,000-page archive cannot drown.
- Read groups on non-branded queries when the homepage sits in a group: navigational demand inflates whichever group contains it. The brand split stacks with the page filter.
Why your group numbers look off
- Groups do not sum to the site total? Expected, twice over: overlapping patterns double-count pages, and anonymized queries drop out of filtered views entirely. Partition cleanly and compare groups to their own history.
- A group suddenly jumps? Check for a URL structure change before celebrating: pages moving into the pattern reads as growth. Migrations and slug changes belong in the annotations for exactly this moment.
- The regex group is bigger than expected? An unescaped special character is over-matching. Retest the pattern against a URL list.
- A group of few pages looks wildly volatile? Small denominators wobble. Read small groups over 90 days, not 28.
Tools that make this easier
- Free, on this site: the GSC regex generator for building and testing group patterns, and the sitemap extractor for pulling the URL list you will define groups against.
- CrawlRaven: topic clusters are content groups maintained for you on the free monitoring dashboard: pages grouped by topic over daily GSC data, cannibalization flagged inside each cluster, and the join attaching crawl health and GA4 engagement to every group trend, so "this section is declining" arrives with its because.
- Honest alternative: SEO Gets has a dedicated content-groups feature over GSC data, well-liked and search-only. A disciplined spreadsheet of saved regexes gets you 80% of either product.
Define the groups this week, save the patterns, and add the portfolio read to your monthly session. The first month usually pays for the setup by itself: somewhere in your blended average, one section has been quietly out-earning the rest, waiting to be noticed.
Frequently asked questions
Does Google Search Console have a content groups feature?
No. As of August 2026 there is no native grouping feature in Search Console. The same analysis is built manually with the Performance report's Page filter: 'URLs containing' for folder-based sections, or a custom regex for groups that do not share a path. Some GSC-layer tools, CrawlRaven and SEO Gets among them, maintain named groups for you on top of the same data.
What are content groups in SEO?
Content groups are named sets of pages analysed as units: a blog section, a product template, a topic cluster, a publish quarter. They exist because per-page data is too granular to show strategy-level patterns and the site-wide total is too blended to show anything. Group-level trends reveal which sections earn growth, which coast, and which decline together.
How do I group pages in Search Console?
Open the Performance report, add a Page filter, and either use 'URLs containing' with a path fragment like /blog/, or choose Custom (regex) for groups that do not share a folder. The filtered view's totals, queries and trends now describe that group alone. Save the patterns somewhere shared and reapply them each session, since GSC does not remember them for you.
What is the best way to define content groups?
Follow the questions you actually ask, not the folder structure. By section when you ask which part of the site performs, by template when you ship template changes, by topic cluster when you invest in topics, by publish era when you changed the content playbook and want to compare output. Most sites need three to six groups; more than ten stops being a portfolio and becomes pages again.
Why do my content group numbers not add up to the site total?
Two reasons. Overlapping patterns count the same page in two groups, so tighten the regexes if you need clean partitions. And on the query dimension, anonymized queries drop out of any filtered view, so even perfectly partitioned groups sum short of the unfiltered property total. Compare each group against its own history and the arithmetic behaves.
Can content groups show why a section is declining?
They show that it declines as a unit, which is the diagnostic half: a whole group falling together points at a shared cause, usually a template, an intent shift on the topic, or a competitor beating the format, rather than at individual pages aging. Confirming which cause requires the page-level and crawl evidence behind the group.
15+ years of growing SaaS websites through SEO | Author, 200-Point Audit Checklist
Aditi has spent 15+ years helping SaaS companies scale organic traffic through technical SEO and content strategy. She is the author of the CrawlRaven 200-Point Audit checklist used by agencies and in-house teams to systematically improve search performance.