The Google Search Console API lets your scripts and workflows pull the same data you see in Search Console (clicks, impressions, click-through rate, and average position by page and query) without clicking through the dashboard. It's free, and for most SEO automation it's the most useful data you have.
If you've ever exported the Performance report to a spreadsheet, hit the 1,000-row limit, and wondered what you were missing, the API is the fix. It returns far more rows, and you can slice by date, page, query, country, and device.
Is this for you?
- You want reports or alerts built on real search data from your own site.
- The dashboard's export limit keeps cutting your data short.
- You want to keep more history than Search Console holds, which is about 16 months.
What you need before you start
- Access to the property in Search Console with the right permissions.
- Either a Google Cloud project with the Search Console API switched on, or a no-code tool that connects through your Google login.
- Your property type: a domain property (which looks like sc-domain:example.com) or a URL-prefix property. Using the wrong one returns empty results.
Your first build: a 28-day decay check
This is the engine behind most “pages that dropped” alerts.
- Request clicks by page for the last 28 days and for the 28 days before.
- Page through the results (up to 25,000 rows per request) until you have everything.
- Match pages across the two periods and work out the change.
- Keep pages with meaningful traffic that dropped by more than a set amount, say 20%.
- For each, request its top queries for both periods to see which searches it lost.
Two things surprise beginners. Data usually runs a couple of days behind, so don't compare “yesterday.” And API totals won't always match the dashboard exactly, because some rare queries are hidden for privacy.
Mistakes to avoid
- Average position is an average. A page at #3 for one query and #40 for another shows a meaningless middle number. Look at it query by query.
- Quotas exist. Batch your requests and store results rather than asking again.
- Save your own history if you want more than about 16 months.
- Double-check the property type if you get empty results.
Where to go next
- Automate Google Search Console: alerts, reports, and to-do lists
- Content decay workflow: the full workflow built on this data
- Automate SEO reporting: reports that explain what changed
- The SEO automation loop: the observe, decide, act, measure cycle every build fits into.
- The SEO autonomy dial: how much control to hand over, and when.
- Examples: what other people have built.