For practical SERP analysis, collect the search results for one query, review the pages and questions Google shows, then use that evidence to decide what your content should offer. Tab Automation’s Google Search workflow handles the collection and optional source-page capture. You make the editorial decisions.
A SERP is a search engine results page. Analyzing it helps you examine the content formats, competing pages, and questions surrounding a query before commissioning an article or refreshing an existing page. Similarweb’s SERP analysis guide describes this broader research process; this guide focuses on collecting and using the evidence in Chrome.
For a freelance SEO marketer, the practical benefit is less copying between tabs and a research document you can revisit or hand to a writer. For an in-house marketer, the same export can support a discussion about what a new page should cover and where it can contribute something useful.
We’ll use “how to choose running shoes” as an illustrative query. The brief below is a template, not a report of a live search or a claim about which pages currently rank.
Collect one search with a clear purpose
Start with a decision: “What would help a first-time runner choose a pair of shoes?” That gives you a reason to inspect the results beyond copying the headings of competing articles.
- Open desktop
www.google.comin Chrome and search forhow to choose running shoes. Use a results page with English headings. - Open the Tab Automation side panel, go to Library, and choose Google Search. If activation is required, follow the Launch License instructions.
- Configure the collection options below before choosing Play. Play uses the current Search results tab; you enter the query in Google yourself.
Keep Chrome open during the run. Start with one search page by leaving Continue to the next page off. A small first pass makes it easier to check the captured evidence before expanding your research.
Choose what to capture
For this content-planning pass, keep Collect Organic Search Results, Scrape Organic Search Result pages, and Save readable page text on. These defaults collect primary organic results and the readable content available on their linked pages. Leave Save raw page HTML off unless your work specifically needs the markup.
Keep Collect People Also Ask Questions and Open each question on to collect questions with expanded answers and sources. Keep AI Overview, related-search, and video collection enabled to inspect those features when present. Leave the separate AI Overview, People Also Ask, and video source-scraping switches off for this first pass; enable one when you also need its linked pages’ text. Turn Collect Nearby Places off for this example.
These are the Google search results and features selected for this pass. Source-page text is captured separately from the pages they link to when the corresponding Scrape option and Save readable page text are enabled. Use each result or feature to answer a specific research question:
| Search result or feature | What to examine for your brief |
|---|---|
| Organic Search Results | Which pages appear as primary organic links? Read their captured source-page text to examine how competitors structure and explain the topic. |
| People Also Ask Questions | Which questions might your reader need answered? Check expanded answers against their sources. |
| People Also Search For | Which adjacent queries deserve a separate manual search? The workflow collects these queries without following them. |
| Videos | Would a demonstration help explain something that is difficult to describe in text? Video-page capture does not provide transcripts. |
| AI Overview | What does Google’s generated answer emphasize, and which sources does it cite? Treat it as a search observation to verify. |
You do not need every switch for every project. For a local-business brief, Nearby Places may be useful context; for an informational article, competing guides and reader questions may be the priority.
Review the evidence in Results
Choose Play and let the pass finish. If a captcha pauses the workflow, solve it in the tab and choose Resume.
Open Dashboard from the side-panel header, choose Datasets, and open the Google Search dataset for your run. Review the result types, source URLs, ranks, and capture statuses. Use Preview where available to read retained source content.
Check coverage before drawing conclusions. A collected result can have a source page whose capture failed or was disabled. Its snippet is not a substitute for having read the full page. Inspect failures and record any research you still need to do manually.
The workflow stores organic and video positions separately, with their search page. Organic rank 3 on page 2 means the third detected primary organic result on that page. It is not automatically “position 13,” and it is not a visibility score across all search features.
One useful detail when reviewing the export: if the same source destination appears in several collected blocks, the workflow captures it once per run and links that content to the referring results. Repeated references to one page should not become several independent pieces of evidence in your brief.
Turn SERP analysis into a content brief
Read the competing pages before choosing an outline. Look for the task each page helps someone complete, the evidence it supplies, and the questions it leaves unresolved. A topic appearing repeatedly is a reason to investigate it, not an instruction to reproduce every competitor’s section.
Copy this table into your brief and replace the prompts with findings from your own run:
| Brief field | What to record |
|---|---|
| Audience | For this illustrative assignment: a first-time runner choosing shoes. Confirm that this matches your business’s intended reader. |
| Likely intent | Your interpretation of the task behind the query, supported by the result types and pages you reviewed. |
| Page format | The format you propose, such as a decision guide, and why it suits the reader’s task. Record what you actually observed before making the choice. |
| Questions to answer | Relevant collected questions, their source references, and any additional questions from your customers. |
| Competing sources | URLs, search page and organic rank, capture time, and notes on each page’s useful evidence or omissions. |
| Original contribution | Something your team can substantiate: for example, a documented fitting demonstration or an interview with a qualified specialist. Assign an owner to produce it. |
Separate observation from interpretation in your notes. “This captured page includes a fitting checklist” is an observation you can link to. “Our readers would benefit from a printable checklist” is your editorial proposal. Keeping both makes a writer’s handoff easier to understand and challenge.
For an existing article, use the same brief to identify specific updates: an unanswered reader question, an explanation that needs evidence, or an example your team could improve. The workflow supplies research material; it does not identify content gaps or write the brief for you.
Export research for your next handoff
From the dataset, choose CSV for spreadsheet review or MARKDOWN for a readable research document. Exports are disabled while the run is running; finish it or pause it before exporting the saved results.
- CSV preserves structured result data for sorting and annotating in a spreadsheet. Import the downloaded file yourself; there is no direct Google Sheets connection or continuing synchronization.
- MARKDOWN includes run metadata and coverage, a ranked result index, attributed search observations, and a source corpus containing each captured page once. Source citations connect observations to the captured text. Attach your brief separately so the writer can distinguish evidence from your recommendations.
- JSON supports further processing. ZIP bundles JSON, CSV, Markdown, and referenced raw HTML files when HTML was captured.
The Markdown export does not call a model or generate conclusions. Its value is a traceable research handoff: the writer can follow a reference instead of guessing which page a note came from.
Know what the collection covers
Organic collection defaults to all detected primary results on the page. Limit results can cap that at 1–20 per page. Pagination supports up to five pages, and a run retains at most 120 result blocks. Collect more questions can expand collection by up to eight extra People Also Ask questions. Review any collection-limit notices before calling a dataset complete.
This workflow supports desktop www.google.com Search with English headings. Other Google domains, Images, ads, and consent bypass are outside its coverage. Source text covers the readable content available in the loaded page; it does not retrieve hidden content or bypass paywalls. Unsupported PDFs and failed pages remain explicit capture errors.
Treat a run as evidence from the search you inspected. It does not supply search volume, keyword difficulty, or a ranking guarantee. Use your own performance data and business knowledge alongside it when deciding what deserves investment.
Results and source content stay local in this browser. Retained earlier runs remain available for review and export after reruns or automation removal. Explicit deletion in Results removes the selected retained content; uninstalling the extension or clearing its storage also removes it. Download research you want to keep independently or share with a client.
For your first pass, aim for one useful deliverable: a brief with a clear reader, a justified format, source-backed questions, and an original contribution your team can produce. The Google Search workflow reference covers the controls when you’re ready to collect the next query.
Editorial note: This guide was drafted with AI assistance and checked against the extension source. The running-shoes query and brief are illustrative; no live search experiment is claimed.