How to Set Up Custom Reports in Google Search Console

Google Search Console custom reports are not a single feature. Search Console has no report builder, so you create a custom report by combining filters, comparisons, and regex inside the Performance report, then moving to Looker Studio, the API, or BigQuery when the interface runs out of room. That ceiling arrives quickly: the Search Console interface exports a maximum of 1,000 rows, while the Search Analytics API and the Looker Studio connector can reach up to 50,000 rows per day per site per search type.

This guide shows you how to build each kind of custom report, which method suits which job, and the data limits that quietly distort results if you ignore them.

What Are Google Search Console Custom Reports?

Google Search Console custom reports are tailored views of your search performance data, built by filtering, comparing, or exporting the Performance report so it answers a specific question, such as “which question queries lose clicks on mobile?” Search Console offers no drag-and-drop builder, so customization happens through filters, regex, comparisons, and external tools.

There are four practical routes, each with different limits:

Start at the top and move down only when a question cannot be answered where you are.

How Do You Build a Custom Report in the Performance Report?

Open Performance, choose Search results, and click + Add filter to combine dimensions such as query, page, country, and device. Add a comparison to evaluate two periods or segments side by side. Together, filters and comparisons create a custom view without leaving Search Console.

Google’s own help documentation confirms you can filter by multiple dimensions at once. The steps are:

1. Click + Add filter and choose Queries or Pages.
2. Choose “containing” for one substring, or Custom (regex) for multiple patterns.
3. Add country, device, or search appearance filters as needed.
4. Add a Compare filter to evaluate two date ranges or two segments.
5. Choose your metrics, group by Queries or Pages, and export from the table.

How do regex filters work in Search Console?

Regex filters let one expression match many queries or URLs. Google uses RE2 syntax, matches are partial by default, and matching is not case-sensitive unless you start the expression with (?-i). Google’s advanced filtering guide recommends separating multiple values with a vertical bar and wrapping the whole expression in parentheses.

Useful starting patterns, which you should test against your own data:

Question and long-query filters are especially useful for understanding how people phrase searches in the era of AI-assisted results. The AI search optimization services overview covers why that conversational phrasing matters for answer engines.

Using the branded queries filter

Since late 2025, you can skip hand-written brand regex in many cases. Google’s branded queries filter segments query data into branded and non-branded views in the Search results Performance report, and Google states it has no effect on how Search ranking works. It is a rollout-based feature, so confirm it appears in your property. Because Google classifies brands automatically, spot-check the results, and keep a regex fallback for edge cases like misspellings of niche product names.

Adding annotations and AI-powered configuration

Annotations let you mark events, such as a site migration or campaign launch, on the performance chart, so later traffic changes have context. Pairing them with dates from Search Savvy’s Google algorithm update tracker makes it easier to separate your own changes from search-wide volatility.

Google is also testing AI-powered configuration, an experimental feature where you describe a report in plain language and the tool suggests filters, comparisons, and metrics. Per Google’s documentation, it is visible only to a small percentage of users, is limited to 20 requests per day, works only for Search results, and cannot export data or sort tables. Always review the filters it applies, since Google warns it may misread a request.

How Do You Connect Search Console to Looker Studio?

Create a new Looker Studio report, add the Google Search Console connector, choose your property, select the table and search type, then build charts from the fields. This turns the Performance data into a shareable dashboard that refreshes automatically, without manual exports.

Practical setup tips:

• Pick the right table for the question. Google distinguishes data aggregated by property from data aggregated by URL, and totals can differ between them.
• Use calculated fields for segmentation. Looker Studio’s CASE and REGEXP_MATCH functions can group queries into branded, informational, and transactional buckets.
• Build around decisions. A single page listing declining pages, rising queries, and device splits beats a dashboard with 30 charts nobody reads.
• Remember the limits. The connector shares the 50,000-row daily cap with the API, which can truncate long-tail query data on large sites.

For query analysis, group queries by intent before charting. Teams that do this often find the same topic spread across several pages, a finding that feeds directly into content planning, the kind of work covered under content strategy and topical authority services.

How Do You Use the Search Console API for Custom Reports?

Send a POST request to the Search Analytics query endpoint with your date range, dimensions, filters, and a rowLimit. The API returns 1,000 rows by default, up to 25,000 per request, and you page through larger results with startRow. It suits automation, scheduled pulls, and datasets too big for a dashboard.

A minimal request body looks like this:

{

  “startDate”: “2026-09-01”,

  “endDate”: “2026-09-30”,

  “dimensions”: [“query”, “page”],

  “dimensionFilterGroups”: [{

    “filters”: [{

      “dimension”: “query”,

      “operator”: “includingRegex”,

      “expression”: “^(how|what|why)\\s”

    }]

  }],

  “rowLimit”: 25000,

  “startRow”: 0

}

Google added includingRegex and excludingRegex operators for the query and page dimensions in October 2021, so the same logic you use in the interface works in code.

One common misunderstanding: the 50,000-row limit applies per day, per site, per search type, not to your whole request. A 30-day pull can return far more rows in total, though adding both query and page dimensions fills the daily cap faster. Check Google’s current documentation before building a pipeline, because quotas can change.

When Should You Use Bulk Data Export to BigQuery?

Use bulk export when you need complete, ongoing, queryable data beyond API and dashboard limits. Search Console sends a daily export to your BigQuery project, and only property owners can configure it.

The export creates searchdata_site_impression (aggregated by property), searchdata_url_impression (aggregated by URL), and an ExportLog table. Per Google’s table reference, data arrives once per day, and rows can repeat keys, so aggregate in your queries rather than assuming one row per query. Setup requires enabling the BigQuery API and granting access to the Search Console service account, which usually needs a developer. Because exports start only from when you configure them, turn this on early.

A Step-by-Step Playbook for Your First Custom Report

1. Write the question first. For example: “Which informational pages lost clicks after August?”
2. Choose the lightest method. Try filters before dashboards, and dashboards before code.
3. Segment the data. Apply page, query, device, and country filters that match the question.
4. Compare, do not eyeball. Use weekly or monthly granularity, which Google suggests to neutralize day-of-week effects.
5. Annotate changes. Record launches, migrations, and known updates.
6. Export and document. Save regex patterns and filter logic so colleagues can reproduce the report.
7. Review monthly. Prune charts that never change a decision.

Common Mistakes to Avoid

• Trusting totals after filtering. Google notes that filtering by query or URL can shift totals because of data truncation and anonymized queries.
• Comparing mismatched dates. A Monday against a Saturday inflates differences.
• Ignoring the row cap. Large sites can lose long-tail rows without any warning.
• Over-relying on AI configuration. It is experimental and can misinterpret prompts.
• Treating branded classification as perfect. Verify it against queries you know.
• Building dashboards nobody uses. A report without an owner and a decision attached is decoration.

If you are unsure what to prioritize once the data is flowing, a keyword and intent review is a sensible next step. Search Savvy’s keyword research services can help translate query patterns into a plan, and the SEO glossary explains unfamiliar terms along the way.

Frequently Asked Questions

Can you create custom reports in Google Search Console?

Not with a dedicated builder. You customize reports using filters, comparisons, and regex in the Performance report, or build fuller reports in Looker Studio, the API, or BigQuery.

How many rows can you export from Google Search Console?

The interface exports up to 1,000 rows. The API and Looker Studio connector can reach up to 50,000 rows per day per site per search type, per Google’s documentation, and bulk export to BigQuery removes most practical limits.

Does Search Console support regex?

Yes. Performance report filters for queries and pages accept RE2 regex, with matches and does-not-match options, and the API supports includingRegex and excludingRegex.

Is Looker Studio better than the API for Search Console reports?

Looker Studio suits dashboards and non-technical teams, while the API suits automation and larger pulls. Both share the same daily row cap, so the choice mostly depends on who maintains the report.

Why do my filtered totals not match the overall totals?

Filtering by query or page can change totals because of data truncation and anonymized queries that Search Console omits. Check totals without query or page filters when accuracy matters.

Is Search Console data exported to BigQuery forever?

Tables are retained indefinitely by default, subject to your Google Cloud project settings. Export begins from setup, though, so historical data before that date will not appear.

Bottom Line

The right custom report is the smallest one that answers your question. Start with filters and regex, add annotations so changes have context, and move to Looker Studio, the API, or BigQuery only when you hit a limit. Document your filter logic, and treat totals with caution whenever query or page filters are applied.

For your next step, pick one decision you need to make this month, build a single filtered view around it, and then decide whether it deserves a dashboard. If you want help turning query data into priorities, Search Savvy’s free SEO tools and service pages are a practical place to continue.