Search Console automation checks your Google Search Console data at set times. It can collect results each week, compare periods, find important changes, and save a report.
A useful first task is a weekly SEO audit. The automation checks clicks, impressions, pages, and search queries. It compares the results with an earlier period. It then checks the actions from the previous report.
This gives you three answers: What changed? What needs attention? What happened after the last change to the site?
This guide explains what to automate and how to compare results. For the detailed setup, use our weekly Search Console audit guide.
What can you automate in Search Console?
Start with a small set of checks that help you make decisions.
| Task | What the automation checks | Useful output |
|---|---|---|
| Weekly performance audit | Clicks, impressions, CTR, and average position | A comparison with the previous period |
| Page review | Pages with the largest click gains and losses | A short list of pages to inspect |
| Query review | Search terms that gain or lose traffic | Queries and their related pages |
| CTR review | Pages with many impressions and fewer clicks | Pages that need a closer check |
| Content update review | Results before and after a recorded edit | Progress against the expected result |
| Index status check | Google's stored index status for selected URLs | Important pages that need investigation |
CTR means click-through rate. Calculate it as clicks divided by impressions, multiplied by 100.
The Search Console API provides Search Analytics, Sitemaps, Sites, and URL Inspection services. It does not provide every report shown in the browser. Full Page Indexing and Core Web Vitals reports need separate checks.
The URL Inspection API checks the version stored in Google's index. It does not test the live page or request indexing. Use it for selected URLs when the audit finds a reason to investigate.
Choose how the automation will run
An automation needs a data connection, a schedule, comparison rules, and a place to save results.
| Method | Use it when | Setup to complete |
|---|---|---|
| SearchConsole.ai with an AI agent | You want to ask questions and review results in plain language | Connect the property and configure a scheduler that can use the connection |
| A workflow tool such as n8n | You want to run requests, calculations, and reports on a schedule | Configure authentication, the schedule, and each step |
| A script with the Search Console API | You need specific comparison rules or report formats | Write the data requests and calculations, then schedule the script |
| BigQuery bulk export | You need stored data for larger analyses | Configure the export, then add queries and report delivery |
For example, n8n's Schedule Trigger can start a workflow each week. Google's BigQuery bulk export can export Search Console performance data each day.
SearchConsole.ai gives an AI agent read-only access to your Search Console data through MCP. MCP is a connection standard for AI tools. The agent or another scheduler must start each scheduled task. The connection alone does not create a schedule.
To use this method, connect SearchConsole.ai, then follow the MCP setup guide. Test one report before you schedule it. Confirm that the scheduled task can use the same property and save its output.
Set up a weekly Search Console audit
Use the same procedure each week. Start with one property and Web search results.
1. Select complete reporting periods
Compare the latest seven complete days with the previous seven days. Also compare the latest 28 complete days with the previous 28 days.
The weekly comparison shows recent changes. The 28-day comparison helps you check whether the change continues over a longer period. Both periods contain the same number of each weekday.
Recent Search Console data can be incomplete. The API supports dataState: final for finalized data. With dataState: all and the date dimension, first_incomplete_date can identify the start of incomplete data. End the comparison before that date. See Google's Search Analytics reference.
Do not assume that yesterday is complete. Confirm data availability for both periods before you calculate changes.
For example, if September 13 is the latest complete date, use these periods:
| Comparison | Current period | Previous period |
|---|---|---|
| Seven days | September 7–13, 2026 | August 31–September 6, 2026 |
| 28 days | August 17–September 13, 2026 | July 20–August 16, 2026 |
Record the exact dates in every report. Search Analytics dates use Pacific Time, even if the scheduled task runs in another time zone.
2. Keep the comparison settings the same
Use the same property, search type, filters, and method of grouping data for both periods. Keep separate comparisons for Web, Image, Video, News, and Discover data.
Compare the same page with its earlier results. Compare the same query with its earlier results. Two lists of top queries can contain different terms.
For a seasonal business, also check the corresponding period from the previous year when data is available. Record differences in holidays or events.
Google's comparison guide explains date, page, country, and device comparisons.
3. Collect totals, then find the affected pages
Start with property totals for clicks, impressions, CTR, and average position. Request these totals separately from the query list.
Then compare pages, queries, devices, and countries. Sort page changes by the number of clicks gained or lost. Include important pages even if they are absent from the latest top-page list.
The API returns top rows and does not guarantee every row. An absent query is not proof of zero clicks. Record retrieval limits and treat an unmatched row as unconfirmed. See the API reference and our data limits guide.
4. Review the largest changes
For each important change, identify the affected page or query. Show the earlier value, current value, and difference.
Use the other metrics to select the next check:
| Observed change | What to check next |
|---|---|
| Clicks and impressions decrease | Affected queries, search demand, and index status |
| Clicks decrease while impressions stay similar | CTR, query mix, device mix, and search results |
| Impressions increase while CTR decreases | New queries and their average positions |
| One page loses clicks while another gains them | Whether Google shows a different page for the same queries |
These checks suggest possible explanations. The report must separate observed results from an unconfirmed cause.
5. Save the report and next actions
Save each report with its period end date. Include the property, settings, source data, and any missing evidence.
Choose up to three actions. Give each action an owner, an affected URL, and a review date. Keep earlier reports so the next audit can check progress.
How to compare results correctly
The following example uses invented values for one page. It does not show a customer result.
| Metric | Previous seven days | Current seven days | Change |
|---|---|---|---|
| Clicks | 1,000 | 820 | −180, or −18% |
| Impressions | 25,000 | 25,000 | 0% |
| CTR | 4.00% | 3.28% | −0.72 percentage points |
| Average position | 8.0 | 8.2 | 0.2 positions worse |
Use these calculations:
Click change = current clicks - previous clicks
Click change % = (click change / previous clicks) × 100
CTR % = (total clicks / total impressions) × 100
CTR change in percentage points = current CTR % - previous CTR %
If the earlier click count is zero, the percentage change is undefined. Show the counts and state that a percentage comparison is unavailable.
Calculate combined CTR from total clicks and impressions. If impressions are zero, mark CTR as unavailable. Do not average the CTR percentages of individual rows. Use the average position supplied for the same page or property. A smaller position number means a higher position.
In this example, impressions stayed the same and clicks decreased. Check the queries and devices that contributed to the loss. Review the page title and search results as possible explanations.
The figures alone do not prove that the title caused the decrease. Also check the longer comparison before you make a change.
Check whether earlier changes helped
A weekly audit can return to actions from earlier reports. Record each change when you make it.
| Field | Example |
|---|---|
| Page | /guides/example |
| Change | Added steps that readers needed to complete the task |
| Publication date | September 1, 2026 |
| Expected result | More clicks from a defined group of queries |
| Reference period | 28 complete days before publication |
| Review rule | Review after the selected period is complete; extend it if volume is too low |
Choose the review period before you inspect the result. Allow time for Google to process the page change. Use longer periods when the site has little traffic.
Compare the same page and query group before and after the change. Check similar unchanged pages for context. A wider traffic increase can affect both groups.
Use one of four findings: improved, declined, no clear change, or too little data. Include the evidence and the next review date.
Google explains that other events can affect performance after a page edit. A before-and-after comparison cannot prove the cause by itself. See Google's guidance on measuring page changes and our content update measurement guide.
Set rules for useful alerts
Use both a percentage change and a minimum click change. For example, you might investigate a weekly loss of at least 20% and 100 clicks.
This is an example rule, not a Google recommendation. Adjust it to the site's normal traffic and the importance of each page.
A decrease from five clicks to three is 40%, but it is only two clicks. A small site may need a longer comparison period before a pattern is clear.
Treat missing data, expired authentication, and failed requests as reporting problems. Do not convert them into zero traffic. Check Google's data anomalies page when a change has no clear explanation.
Save a report each week. Send an alert when a defined rule is met, a scheduled review is due, or the task fails.
A prompt for your weekly automation
Replace the example property and output location. Give the prompt to a scheduled agent that can access the data and earlier reports.
Run a weekly Search Console audit for sc-domain:example.com.
Save reports in: [report folder].
Use Web search data. Confirm the latest complete reporting date.
Compare seven complete days with the previous seven days.
Also compare 28 complete days with the previous 28 days.
Record exact dates, filters, and data limits.
Use the same settings for each comparison.
Report property totals for clicks, impressions, CTR, and position.
Calculate CTR from total clicks and impressions.
List the largest page and query click gains and losses.
Show counts and percentage changes where they are defined.
Check device and country differences for important changes.
Do not treat missing rows or failed requests as zero traffic.
Read the previous report and the record of site changes.
Check earlier actions whose review dates have arrived.
Separate observed results from possible causes.
State when there is too little data for a conclusion.
Recommend up to three actions, each with a reason and review date.
Report recommendations without editing the website.
Save the report in the selected folder with its period end date.
Preserve earlier reports and link to the previous report.
Confirm that the report was saved.
Report a failed run separately from a performance change.
Use the weekly audit setup guide to add the schedule, file structure, and alert rules.
Start with one repeatable task
Connect one property. Run the weekly audit once. Check its dates, calculations, and recommendations against the source data.
Then schedule the same task and review the first saved report. Add further checks when the weekly comparison gives you useful, consistent results.