How to capture SERP screenshots with a logged-out browser profile

To capture a search engine results page (SERP) without an account session, open a fresh private browser window, do not sign in, enter the exact query, and save a screenshot with a unique filename. Record the time, location, language, browser, device, and capture scope alongside the image. A logged-out capture documents what appeared in that session; it is not a universal or context-free version of the results.

1. Start a fresh, signed-out browser session

In Chrome on desktop, choose More → New Incognito window. The keyboard shortcut is Ctrl+Shift+N on Windows, Linux, and ChromeOS, or Command+Shift+N on Mac. Chrome does not automatically sign you into Google or other websites in Incognito. Do not sign into Google, Microsoft, or another search account during the capture.

Check whether browser management, extensions, a proxy, or a VPN changes the environment. Private browsing limits local retention, but it does not make you invisible: sites and network operators such as an employer, school, or internet provider may still observe activity. Incognito keeps cookies and site data for the duration of the private session so websites can function; it removes them after all Incognito windows close. Bookmarks and downloaded files remain on the device. See Chrome’s Incognito documentation.

2. Search the exact query and note its context

Type the query exactly as intended, preserving spelling, punctuation, and wording. Before capturing, note the visible location, language, and any Google Search footer or customization indication about personalized results. Google’s Search settings and help labels may change, so check the current interface rather than relying on an old set of steps.

Signed-out does not mean context-free. Google documents time, location, language, device type, recent searches, and personalization as factors that can produce different results. Signed-out autocomplete can also use recent searches. Even when personalization is disabled, the search still uses query context. Google notes: “Search still uses the context of your query to improve results.” See Why your Google Search results differ from others and Google’s Search context and customization information.

Use the same search engine, query, location or VPN endpoint, language, account state, device, and viewport when comparing captures. If the purpose is to study one variable, change that variable deliberately and hold the others as constant as practical.

3. Capture the results page in Firefox

Firefox Developer Tools include a console :screenshot helper. Open the SERP in Firefox, open Developer Tools, select the Web Console, and run a command such as:

:screenshot --fullpage --filename="google-running-shoes-us-2026-10-04-001.png"

This captures the full page. For a viewport screenshot, omit --fullpage:

:screenshot --filename="google-running-shoes-us-2026-10-04-001.png"

Firefox’s helper supports options including filename, full-page capture, selector, delay, and device pixel ratio. Consult Firefox’s screenshot documentation for the current syntax and available options. A selector capture is useful when a specific element is the subject, but a SERP comparison usually needs the search results area and enough surrounding context to identify the page.

You can also use your operating system’s screen capture controls for a visible viewport. The exact controls depend on the OS and are not covered here; verify that the saved image shows the intended browser window and does not include unrelated content.

4. Save and review the image

Give each capture a unique, descriptive filename. Include a short query slug, search engine, locale, date, time or sequence number, and whether it is full-page where helpful. For example:

google-running-shoes-us-2026-10-04-1430-fullpage-001.png

Firefox warns that another capture to the same filename overwrites the earlier image. After saving, open the file and verify that it is the expected SERP, that the intended viewport or page length is included, and that the capture is legible. Keep the original image unchanged if it is being used as research evidence; make annotations on a copy.

5. Attach reproducibility metadata

Store a small text file, spreadsheet row, or record next to each screenshot. This turns an image into an interpretable observation and makes later comparisons more useful.

Capture: timestamp with timezone, filename, and whether it is viewport-only or full-page.
Search: engine, exact query, and visible personalization or Search customization state.
Context: country or region, VPN endpoint if used, interface language, and any relevant location setting.
Environment: browser and version, operating system, device, viewport dimensions, and device pixel ratio if relevant.
Account state: signed out, and whether the capture used a fresh private session.

These details matter because time, location, language, device, and personalization can affect results. They also reflect the sort of environment information recorded in a published logged-out SERP research workflow. A 2025 Dutch parliamentary report, for example, describes Google and Bing captures from 25–26 November 2024 using Firefox on Windows, a Missouri VPN endpoint, and no Google or Microsoft account login. Some generated search components still appeared while logged out. See the reported research methodology.

What a logged-out SERP screenshot can establish

A screenshot is evidence of what appeared in one browser session at one time, for one query and context. It does not prove that every user saw the same ranking or page. Google says most searches from the same location return the same results, while some content can be personalized. Search content changes frequently, and data-center rollouts can differ.

A 2017 study, Measuring Personalization of Web Search, reported that 11.7% of Google Web Search results and 15.8% of Bing Search results differed due to personalization on average in its study. The authors found variation by query and result rank; the measured personalization was associated with signed-in accounts and searcher IP address. Treat those percentages as findings from that study, not current platform-wide rates. See Hannak et al., 2017.

Comparing two or more SERP captures

Write down the question the comparison should answer, such as whether results differ by location or account state.
Capture each condition with the same engine, exact query, language, browser, OS, device, viewport, and capture scope where practical.
Change only the variable under investigation. If that is not possible, record each difference.
Capture close together in time and include timezone. Search results can change as content and data-center rollouts change.
Compare the images alongside their metadata. Distinguish result changes from layout, viewport, or generated-component changes.

For a location comparison, for example, keep the query and browser setup the same while changing the location or VPN endpoint. For a signed-in versus signed-out comparison, record the account state explicitly and avoid describing either capture as a universal baseline.

Common problems and fixes

Symptom	Likely cause	Fix
The SERP still looks personalized.	Signed-out search can still use location, language, device, time, recent searches, and other context; a browser or network may also be managed.	Confirm the account state, inspect visible Google customization information, record location and language, and disclose any management, VPN, or proxy setup.
Two logged-out screenshots show different results.	Time, location, language, device, query wording, network, and result updates can differ.	Compare the metadata, repeat captures close together, and state which variables were held constant.
The screenshot file is missing or the old capture replaced.	The command may have saved to a different location, or the same filename was reused and overwritten.	Use a unique filename and inspect the browser’s download/save location after each capture.
The screenshot is cut off or too long to inspect.	The capture scope was viewport-only or full-page when the other was intended.	Use or remove Firefox’s `--fullpage` option as appropriate, then reopen the image to confirm its dimensions and contents.
The Firefox console rejects the command.	The helper syntax may differ from the current Firefox version, or the command was entered outside the Web Console.	Open the Web Console and check the current Firefox screenshot documentation for option names and syntax.
A private window is signed into an account.	The session was manually signed in, or a managed environment changes browser behavior.	Close all private windows, open a new Incognito session, and verify the search page is signed out before capture.
The screenshot includes unrelated desktop content.	An OS-level capture selected the wrong region or window.	Recapture the browser window or use Firefox’s page screenshot helper, then review the saved image.

Performance, reliability, and cost

The browser method is free and needs no capture service. Its main costs are manual setup, review, and maintaining consistent metadata. Full-page captures can create tall images that are harder to compare and share; use viewport capture when the question concerns only the initially visible results, and full-page capture when lower results or page content matter. Record that choice.

For repeatable research, use a fresh private session for each planned capture, preserve unique filenames, and record the environment immediately. This reduces accidental carry-over and makes failures easier to diagnose, though it cannot freeze changing search results or remove contextual differences. Do not treat a failed or incomplete capture as a valid result; repeat it and document the retry.

Or skip the browser setup

ScreenshotNeo is a website screenshot API and MCP server. Its one-call API can capture a URL as an image or PDF; see the API documentation for parameters and response details.

curl -G "https://api.screenshotneo.com/v1/shot" -d access_key=YOUR_API_KEY --data-urlencode url=https://www.google.com/search?q=running+shoes -o serp.webp

For a logged-out SERP, make the URL represent the exact query you want to capture. Search engines may redirect or alter the page based on context, so preserve the response and record the capture time and relevant URL/context. ScreenshotNeo removes cookie banners, popups, and chat widgets before the shot; each cleanup step can be turned off. Bot checks, blank pages, failed loads, timeouts, and cache hits are never billed, and response headers report the page verdict and billing status. An MCP server lets AI agents use take_screenshot, get_page_info, and capture_pdf. The free plan includes 1,000 screenshots a month with no card; paid plans start at $5 for 3,000 screenshots.

Create a free ScreenshotNeo account to get 1,000 screenshots a month with no card.

FAQ

Does Incognito show unbiased Google results?

It avoids automatic account sign-in and starts a separate local session. It does not remove location, language, device, time, or other search context, so call the result a logged-out capture from a specified environment rather than unbiased or universal.

How do I take a Google search screenshot without being logged in?

Open a fresh Incognito window, do not sign in, search the exact query, and capture the page with Firefox’s screenshot helper or your OS capture controls. Record the context with the image.

Should I capture the whole page?

Capture the viewport when only visible results matter. Capture full-page when the study includes results below the fold or the complete SERP layout. State which scope you used.

What should I record so someone can reproduce the screenshot?

At minimum: exact query, engine, timestamp and timezone, location, language, browser and version, OS/device, viewport, signed-out state, visible personalization setting, and viewport or full-page scope.