Google Search Console Regex: Filters, Examples & How to Use Them

Want to filter branded queries, find question-based keywords, or compare blog sections without checking each term individually? Google Search Console regex lets you group search queries and page URLs using simple expressions. In the Performance report, regex filters can include multiple phrases at once or exclude matches such as your brand name. Start with a pattern like seo|search engine optimization, then refine it for your own site.

In this guide, you will learn how to use regex in Google Search Console, copy practical filters for branded and non-branded searches, and test results before reporting on SEO performance. The examples use illustrative strings, not live OneMetrik Search Console data.

What Is a Regex Filter in Google Search Console?

Regex stands for regular expression, a way to describe text patterns. In the Search results Performance report, a regex filter selects Queries (search phrases) or Pages (URLs) that match an expression. It filters the report; it does not change rankings, create keywords, or recover hidden queries.

For example, seo|search engine optimization matches visible queries containing either phrase. Search Console uses RE2 regex syntax, generally matches text anywhere in the string, and ignores case unless you explicitly switch case sensitivity. These rules are documented in Google’s Search Console filtering guide.

How to Use Regex in Google Search Console (Step by Step)

1. Open the verified property in Google Search Console and navigate to Performance → Search results.
2. Choose your Search type (for example, Web), date range, and any country or device filters you need.
3. Select + Add filter, then choose Queries or Pages.
4. Select Custom (regex), paste the expression, and choose Matches regex or Doesn’t match regex.
5. Click Apply. Inspect the chart and the corresponding Queries or Pages table, then export the report if you need to compare it later.

First filter to try: Use Queries → Custom (regex) → Matches regex with the expression below. It groups visible queries mentioning either of two SEO terms.

seo|search engine optimization

Notice that this matches phrases containing seo, not only the exact query seo. It can also match unwanted fragments in other words. For clean segmentation, check sample rows and use boundaries when necessary.

Google Search Console Regex Filter Examples: 8 Patterns to Copy

These are starting points, not universal audience or intent definitions. Paste one expression into the appropriate filter, review the matching rows, and refine it for your site’s real queries.

The case-sensitive example is a syntax demonstration, not usually a useful standalone traffic segment. The regex patterns above work with RE2-compatible features; no lookaheads or backreferences are required.

Practical Examples With Expected Matches

1. Filter Branded and Non-Branded Queries in Google Search Console

To filter brand keywords in Google Search Console, start with the variants that people actually use. For OneMetrik, the example below includes the exact brand name, a space or hyphen variant, and one possible alternative spelling. Remove speculative variants that do not appear in your real query data.

(onemetrik|one[ -]metrik|onemetric)

Matches: one metrik seo, onemetrik google ads. Does not match: seo dashboard template. This works as a Google Search Console branded queries filter with Matches regex selected. To create a visible non-brand segment, use the same pattern with Doesn’t match regex. Check for false positives when a brand name occurs inside a longer unrelated word.

For a stricter brand expression, add boundaries around the name. This version is more selective in English-letter-and-number contexts:

(^|[^a-z0-9])(onemetrik|one[ -]metrik|onemetric)([^a-z0-9]|$)

A practical use is comparing how much discoverable query traffic is branded versus non-branded in an SEO dashboard tied to pipeline. Remember that it is an estimate based on reported queries, not a complete breakdown of all Google searches.

2. Group Question-Led Queries for New Content

^(how|what|why|when|where|which|who|can|does|is|are)\b

Matches: how to track rankings. Does not match: best seo agencies. The ^ means the query must start with one of these words, while \b prevents matching the start of a longer word. It will not catch every informational query; use the matching rows as a list of content opportunities, not proof of informational intent.

Use the matching rows to identify pages you could strengthen, and cross-check ranking positions using our GSC keyword ranking tutorial. For gaps that competitors cover better, follow the SEO competitor analysis workflow.

3. Identify Pricing and Comparison Searches

(^|[^a-z0-9])(vs\.?|versus|compare|comparison|alternatives?)([^a-z0-9]|$)

Matches: ahrefs vs semrush and semrush alternatives. Does not match: seo pricing guide. Use it to find comparison searches, then evaluate whether your existing comparison article or landing page is the best match. If you also want pricing queries, run the separate pricing regex in the cheat sheet.

4. Measure a Blog or Tutorial Directory

https?://[^/]+/blogs/

Matches: https://onemetrik.com/blogs/seo-dashboard/. Does not match: https://onemetrik.com/marketing-tutorials/seo-audit-checklist/. Apply this to Pages, not Queries. It looks for a complete hostname followed by /blogs/. If your website has a different URL structure, replace that directory name.

For the tutorials section of OneMetrik, use https?://[^/]+/marketing-tutorials/. You can then compare performance within a specific content group and use an SEO audit checklist when pages in that group require follow-up.

Regex Syntax and Common Mistakes

Important RE2 limitation: Search Console does not support PCRE lookarounds such as (?!brand), or backreferences such as \1. To exclude brand terms, use the built-in Doesn’t match regex operator instead of negative lookahead. An invalid expression can produce no results.

For accurate filtering, write down the expression, the dimension, the positive or negative operator, and at least one string that should match and one that should not. Check case, spelling, URL paths, and any other active Search Console filters.

How to Validate a Regex Filter Before Reporting Results

1. Test the exact pattern against a known matching query or URL and a known non-matching example in an RE2-compatible tester.
2. Apply the filter in the correct Queries or Pages dimension.
3. Inspect several real rows for false positives. A pricing keyword is not always commercial intent, and a question starter is not always an informational query.
4. Keep Search type, device, country, search appearance, and date range consistent if comparing segments or periods.
5. Capture the Search Console filter settings or include them in an export log so another person can reproduce the report.

To investigate why pages underperform after segmenting the data, use the on-page SEO checklist rather than making changes solely because a filtered query’s clicks fell.

Compare Equal 28-Day Periods Without Misreading the Numbers

Choose two complete, equal 28-day windows, such as 27 August to 23 September 2026 versus 24 September to 21 October 2026, inclusive. Use the same property, Search type, regex, operator, country, device, and other filters. Only compare days whose data is finalized.

For a non-zero baseline, percentage click change is (current clicks - previous clicks) / previous clicks × 100. Calculate impressions separately. Don’t add or naively average row-level CTR and position values. Instead, use the report’s aggregate measurements and interpret them alongside the performance of relevant landing pages.

Why Filtered Totals May Not Add Up

Search Console anonymizes some query strings to protect privacy, and query filters cannot allocate that hidden text to your regex segment. There are also reporting and table-row limits. Consequently, the branded and non-branded filtered totals may not sum to the property’s unfiltered total. Page and query views may also differ because of canonical URL and aggregation rules. Google describes these constraints in its data discrepancy guide.

Google Search Console Regex Filters: API vs Interface

For a one-off investigation, the Performance report is usually easiest. If you need automated exports, the Search Analytics API supports RE2 filtering using dimensionFilterGroups with includingRegex or excludingRegex. The API’s request dates are interpreted in Pacific Time, and each request can ask for up to 25,000 rows. Pagination can retrieve additional eligible rows, but Google’s internal reporting limits mean this is still not a complete query inventory.

If you want an AI assistant to investigate these search segments using authorized API data, see our Google Search Console MCP setup tutorial. It explains connecting Search Console to Claude, using read-only credentials and validating query results.

Here is an illustrative request body that filters queries beginning with question words:

{
  "startDate": "2026-09-24",
  "endDate": "2026-10-21",
  "dimensions": [
    "query"
  ],
  "type": "web",
  "dimensionFilterGroups": [
    {
      "groupType": "and",
      "filters": [
        {
          "dimension": "query",
          "operator": "includingRegex",
          "expression": "^(how|what|why|when|where|which|who|can|does|is|are)\\b"
        }
      ]
    }
  ],
  "rowLimit": 25000
}

When placing an expression inside JSON, escape each regex backslash as \\. This code is an example request, not an API result or a claim about performance on the OneMetrik property.

Make Regex Segments Part of Your SEO Workflow

A useful reporting setup does not need dozens of permanent patterns. Start with branded, non-branded, question-led, comparison, and key URL-directory segments. Save each expression and its definition, review examples during reporting, and connect meaningful changes to content decisions and on-site outcomes. Regex filters make it easier to investigate performance, but they do not establish whether SEO caused pipeline growth on their own.

Frequently Asked Questions