Extract Query by Page Using the Searchconsole API wrapper (Python) - JC Chouinard

Why query by page is worth automating

The Search Console interface is fine for spot checks, but when you need the queries driving every important URL across hundreds of pages, the UI is slow and export limited. The Search Analytics API returns the same data programmatically, and the searchconsole Python wrapper makes it approachable. Pulling query by page lets you find pages ranking for queries their content never mentions, spot cannibalization where two URLs compete for one query, and build content briefs from real demand rather than guesses.

Authenticating and selecting a property

The wrapper handles OAuth using a client configuration you download from a Google Cloud project with the Search Console API enabled, plus a saved credentials file so you do not re authenticate every run. Once authenticated you select one of your verified properties, for example your domain property, and you are ready to query. Keep the client and credential files out of version control, since they grant access to your Search Console data.

Building the query by page report

The core pattern is a fluent chain: start from the property, set a date range, add the dimensions you want, then call get. To break queries down by page, request both the page and query dimensions together. Convert the result to a pandas DataFrame and you can filter, group and export it like any other dataset.

import searchconsole

account = searchconsole.authenticate(
    client_config='client_secrets.json',
    credentials='credentials.json')

webproperty = account['https://example.com/']

report = (webproperty.query
          .range('2026-01-01', '2026-03-31')
          .dimension('page', 'query')
          .get())

df = report.to_dataframe()
top = df.sort_values('clicks', ascending=False)
print(top.head(20))

From that DataFrame you can group by page to count how many distinct queries each URL earns, or group by query to detect two URLs splitting the same demand. For programmatic indexing checks alongside this, pair it with the GSC URL Inspection API.

Limits and gotchas that bite people

Three constraints matter. First, row limits: a single request returns at most 25,000 rows, so for large sites you paginate using the startRow parameter and stitch the pages together. Second, data window: Search Console retains roughly 16 months of data and the most recent two to three days are usually incomplete, so avoid reporting on yesterday. Third, sampling and anonymized queries: rare queries are omitted for privacy, so summed clicks by query will not perfectly equal the property total. Also mind the daily quota on API requests and add simple retry handling for transient errors.

For a fuller pipeline and storage patterns, see our guide to Search Console API data extraction and the Google Search Console glossary entry for definitions.

Frequently asked questions

Source: https://www.jcchouinard.com/searchconsole-api-wrapper-python/