Webpage Diff Checker

🔎 Find webpage changes without checking by hand

Webpage Diff Checker is for SEO teams, developers, and site owners who track public pages. It compares each page with its last good baseline and returns added or removed text, metadata and link changes, severity, and key page facts.

• Website change monitoring checks public pages again and shows what changed.
• Webpage diff checking compares visible text or HTML with the page's last good baseline.
• Competitor page monitoring watches public product, feature, and news pages for changes.
• SEO metadata change tracking finds changes to titles, descriptions, canonical links, hreflang, and robots rules.
• Structured data change monitoring tracks JSON-LD, microdata, RDFa, and Open Graph changes.
• Website link change tracking finds internal and external links that were added or removed.
• Pricing page change monitoring tracks price-page changes with watched words or a selected CSS section.
• Policy page change monitoring keeps separate named baselines for terms and policy pages.

📦 Webpage checks and change details

Each saved dataset row is one page that was fetched and checked. It includes:

• requested and final URL, check time, HTTP status, response time, redirect state, page size, and content fingerprint
• new, unchanged, or changed status
• a short summary, change magnitude, severity, and categories
• added and removed text, plus watched keywords found in the changes
• SEO and semantic metadata changes with before and after values
• internal and external links added or removed

The run summary also gives totals for checked, changed, new, unchanged, and failed pages. Failed pages include error details there. A failed fetch does not replace the last good baseline.

🚀 Compare pages across runs

1. Add one or more public HTTP or HTTPS webpage URLs.
2. Give the monitor a stable name so it keeps its own baselines.
3. Choose visible text or HTML. You can also pick a CSS section or content scope.
4. Add watched keywords or rules that remove known page noise if needed.
5. Choose all page rows or only new and changed pages. Failed checks stay in the run summary.
6. Run the Actor. Run the same setup again to compare with the saved baseline.

The first successful check for a page creates its baseline. Later successful checks compare with it and then save the new good baseline.

🧾 Input

urls is required. Add one or more public HTTP or HTTPS webpages. The schema sets no fixed maximum for this list. Use a different monitorName for each separate set of baselines.

Input fields

Field	Type	What it does
urls	array of objects	Required. Lists the public webpages to check. Add at least one URL.
urls[].url	string	Required URL for one public HTTP or HTTPS webpage.
monitorName	string	Names the baseline set. It defaults to default; use a different name for each separate monitor.
comparisonType	string	Chooses visibleText or html for the content comparison. It defaults to visibleText.
contentScope	string	Chooses wholePage, mainContent, or articleContent. Main content skips common headers, menus, sidebars, and footers. Article content uses the main article or post.
cssSelector	string	Optionally compares only the first page section that matches a CSS selector. Leave it blank to use contentScope.
watchedKeywords	array of strings	Adds words or phrases to find in detected changes.
excludedPageParts	array of strings	Leaves out selected headers, navigation, sidebars, footers, cookie banners, popups, or forms before comparison.
excludedSelectors	array of strings	Leaves out page elements that match the CSS selectors.
excludedPatterns	array of strings	Leaves out changing text that matches the supplied regular expressions.
resultMode	string	all saves every successful page check. changedOnly saves only new or changed pages. Failed checks stay in the run summary. It defaults to all.

This is the public default input used in a successful current-beta run:

{
  "urls": [
    {
      "url": "https://example.com/"
    }
  ],
  "monitorName": "default",
  "comparisonType": "visibleText",
  "contentScope": "wholePage",
  "resultMode": "all"
}

The Actor checks public pages only. It does not log in to private pages or make screenshot or pixel diffs.

📤 Output

Each successful webpage check is saved as one dataset row. All rows use the same shape. status is new for the first successful check, unchanged when the comparison finds no change, or changed when it finds a difference. metadataChanges and linkChanges can be empty when no such changes are found.

Page check fields

Field	Type	What it does
requestedUrl	string	The webpage URL submitted in urls.
finalUrl	string	The webpage URL after redirects.
status	string	Shows new, unchanged, or changed.
checkedAt	string	The date and time of the check.
httpStatus	number	The successful HTTP response status.
responseTimeMs	number	How long the webpage response took, in milliseconds.
contentSizeBytes	number	The fetched webpage size in bytes.
fingerprint	string	A stable fingerprint of the compared content.
change	object	Summarizes content changes and their importance.
change.summary	string	Short plain-text summary of the change.
change.magnitude	number	Share of compared content that changed, from 0 to 1.
change.severity	string	Practical importance: none, low, medium, or high.
change.categories	array of strings	Change kinds: content, seo, semantic, links, or watchedKeyword.
change.addedText	array of strings	Useful text added to the compared content.
change.removedText	array of strings	Useful text removed from the compared content.
change.watchedKeywords	array of strings	Watched words or phrases found in the changes.
metadataChanges	array of objects	SEO and semantic metadata values that changed.
metadataChanges[].kind	string	Metadata group, such as title, description, canonical, hreflang, robots, openGraph, jsonLd, microdata, or rdfa.
metadataChanges[].name	string	Metadata field or item name.
metadataChanges[].before	string or null	Prior value, or null when the metadata was added.
metadataChanges[].after	string or null	Current value, or null when the metadata was removed.
linkChanges	object	Internal and external links added to or removed from the page.
linkChanges.addedInternal	array of strings	Internal links added since the last good baseline.
linkChanges.removedInternal	array of strings	Internal links removed since the last good baseline.
linkChanges.addedExternal	array of strings	External links added since the last good baseline.
linkChanges.removedExternal	array of strings	External links removed since the last good baseline.

Example changed row

This row is copied from a successful current-beta run:

{
  "requestedUrl": "https://www.random.org/strings/?num=1&len=12&digits=on&upperalpha=on&loweralpha=on&unique=on&format=html&rnd=new",
  "finalUrl": "https://www.random.org/strings/?num=1&len=12&digits=on&upperalpha=on&loweralpha=on&unique=on&format=html&rnd=new",
  "status": "changed",
  "checkedAt": "2026-09-10T12:47:37.201Z",
  "httpStatus": 200,
  "responseTimeMs": 1345,
  "contentSizeBytes": 12249,
  "fingerprint": "sha256:1f390f1dc94b3479145a6ade0598eb08662c33eec4b8ca99be69e6aeab56d24c",
  "change": {
    "summary": "Found 1 added and 1 removed text parts.",
    "magnitude": 0.024096385542168676,
    "severity": "medium",
    "categories": [
      "content",
      "watchedKeyword"
    ],
    "addedText": [
      "qenp1r3thOdb"
    ],
    "removedText": [
      "lJghszJ7hElt"
    ],
    "watchedKeywords": [
      "lJghszJ7hElt"
    ]
  },
  "metadataChanges": [],
  "linkChanges": {
    "addedInternal": [],
    "removedInternal": [],
    "addedExternal": [],
    "removedExternal": []
  }
}

Open the dataset for page rows. Open the output key-value store record for run totals and failure details.

💳 Pricing

This Actor uses pay-per-event pricing. One Page check event is charged for each webpage that is successfully fetched and checked for changes. Failed fetches and failure details do not create this event. Check the Pricing tab for current tier prices.

🔌 Integrations

Use Apify datasets, the API, schedules, webhooks, and platform integrations to run repeat checks and send the output to your workflow.

❓ FAQ

What happens on the first run?

The first successful check creates a new baseline. There is no older page content to compare yet.

What happens if a page fails to load?

The run summary lists the failure and its details. The Actor keeps the last good baseline, so a failed page does not become the next comparison source.

How can I avoid cookie banners and changing counters?

Common changing noise is cleaned on its own. You can also leave out common page parts, CSS selectors, or text matched by regex patterns.

Can I watch only one part of a page?

Yes. Add a cssSelector, or choose main content or article content as the content scope.

Can it check private pages or compare screenshots?

No. It checks public HTTP and HTTPS pages without a login. It compares visible text or HTML, not pixels or screenshots.

Can it find old versions from the Internet Archive?

No. It compares repeat checks made with this Actor. Use Wayback Machine Search when you need saved Internet Archive history.

Should I use one monitor name for every task?

Use one stable name for each monitor. Give separate tasks different names so their baselines do not replace one another.

📝 Changelog

v0.0

• Initial release.

🆘 Support

For issues, questions, or feature requests, file a ticket and I'll fix or implement it in less than 24h 🫡

🔗 Related Actors

• Webpage Text Extractor extracts clean text or Markdown from public pages.
• Website URL Crawler & Link Extractor builds the URL list for a site before you monitor chosen pages.
• Wayback Machine Search finds older page snapshots from the Internet Archive.
• Seobility SEO Checker runs a current on-page SEO audit alongside change tracking.
• Sitemap Sniffer finds public sitemap files and page URLs to check.

Made with ❤️ by Maxime Dupré