Visual website monitoring is the practice of regularly capturing a screenshot of a live web page and comparing it, pixel by pixel, against a known-good baseline. When the new screenshot differs beyond a set threshold, you get an alert, so you catch problems a server-status check can never see.
A visual monitor loads your page in a real browser, the same way a visitor would, and saves a full screenshot. After you confirm the known-good state, that screenshot becomes the baseline. On each later scheduled visual check the monitor renders the page again, lines the two images up, and measures the percentage of pixels that changed. A small, expected change (a rotating testimonial, a live counter) stays under the threshold and is ignored. A large change (a hero image that vanished, a layout that collapsed, a page replaced with spam) crosses the threshold and can trigger an incident with both screenshots and the difference highlighted after applicable confirmation rules. Because the check is based on what the page actually looks like, it catches failures that are invisible to traditional uptime monitoring: the server keeps replying normally while real visitors stare at a broken or defaced page.
Uptime, content, and visual checks answer different questions
Uptime monitoring asks whether the configured request succeeded. Content monitoring fetches source HTML and compares visible text extracted from configured selectors. Visual monitoring is the one of these three that compares the rendered picture, by comparing the rendered picture. A site can pass uptime and content checks and still look broken, for example when a stylesheet fails to load: the words are all present in the code and the server responds, but the page renders as a wall of unstyled text. Visual monitoring can flag that on its next scheduled capture. Teams should combine the layers that match their risks without assuming any set catches every possible failure.
Failures that respond with a healthy server but a broken page
- Website defacement, hacked-in spam, and malicious redirects
- Broken or unstyled layouts after a deploy or plugin update
- Missing or broken images, logos, and hero sections
- Failed third-party scripts and widgets (chat, checkout, fonts)
- Error or certificate-warning pages shown to real visitors
- Blank or partially rendered pages
What it looks like in practice
On Room Service Resources (January 2026), a page-builder update finished successfully but failed to rebuild the cache, leaving the hero section as raw, unstyled content for every visitor while the server stayed healthy. Visual Sentinel detected a 48% visual difference on the next check and alerted the team, who fixed it in about ten minutes.
On Flex Town (February 2026), a similar update broke the homepage entirely (the hero image gone, the content collapsed) while the site was technically up. Visual Sentinel flagged an 84% pixel difference, alerted the team, and the team rolled the update back. Both incidents were completely invisible to uptime monitoring. See the full case studies.
Answers, in plain English.
Can't find what you're looking for? Email support.
What is website defacement monitoring?
Defacement monitoring looks for page evidence that may indicate an unwanted visible change. Scheduled screenshot comparison can reveal a large visual difference, while Starter-or-higher content monitoring can compare visible text extracted from configured selectors in fetched source HTML. These signals do not prove that a change was unauthorized, execute page JavaScript, scan malware, or guarantee a fixed detection time.
How does visual regression monitoring work?
A visual monitor loads a public page in a browser, saves a full screenshot as the confirmed baseline, then renders the page again on its visual schedule and measures the percentage of pixels that changed. A change above the configured threshold can open an incident with screenshot and diff evidence after applicable confirmation rules.
How is visual monitoring different from content monitoring?
Content monitoring fetches source HTML and compares visible text extracted from configured CSS selectors. Visual monitoring renders the page and compares screenshots. A page can keep the same text while a stylesheet failure makes it look broken; conversely, a configured price can change without creating a large pixel difference. The two evidence types are complementary.
Does visual monitoring need code or a browser extension?
No browser extension or page-code change is required for a public visual target. Add the URL, configure the capture, and confirm a known-good baseline. Authenticated multi-step login flows are not currently scripted by Visual Sentinel, so use a transaction-monitoring tool when that is required.
Catch the broken page
before your visitors do.
See your site's first check, then try a paid plan for 7 days (card required). Paid plans start at $15/mo.
Run a website of your own? Read about keeping watch over your own website, so an alert tells you when something on it breaks.
See your site checked
Add your site and see its first check. Then try a paid plan for 7 days (card required) or subscribe, from $15 USD a month.
Start 7-day trial