Website Change Intelligence Monitor

Monitor public web pages and detect meaningful content changes between runs instead of repeatedly downloading and comparing pages yourself.

What it detects

• Page title changes
• Meta-description changes
• Visible text changes
• Approximate text-change percentage
• Added / removed link counts
• Added / removed text samples
• Word-count changes
• Final URL changes after redirects
• CHANGED, MAJOR_CHANGE, or UNCHANGED status

The first run creates a baseline, so existing page content is not incorrectly reported as a new change.

Use cases

• Competitor pricing-page monitoring
• Product-page change tracking
• Terms / policy monitoring
• Press / newsroom monitoring
• Documentation monitoring
• Landing-page monitoring
• Partnership / careers page monitoring
• Public company website intelligence

How it works

1. Fetches each public page
2. Removes scripts, styles, and other non-content elements
3. Optionally monitors only a CSS-selected section
4. Normalizes visible text and links
5. Compares with the previous successful snapshot
6. Emits a concise summary plus detailed change records
7. Saves the new snapshot for the next run

Input

• Up to 50 public HTTP/HTTPS URLs
• Stable Monitor ID
• Optional CSS content selector
• Optional CSS selectors to ignore
• Link tracking toggle
• Major-change threshold
• Text-size safety cap
• Optional baseline-detail output

Output

Each page receives a summary with:

• Page title
• Final URL
• HTTP status
• Change signal
• Text-change percentage
• Word-count delta
• Added / removed link counts
• Snapshot hash
• Observation timestamp

When a page changes, a PAGE_CHANGED record also includes short added/removed text samples and changed metadata fields.

Scheduling

Daily or every 6–12 hours is usually enough for business monitoring.

Pricing

Recommended Store launch pricing:

• $0.20 per Actor start
• No additional per-result charge

There is no paid upstream API or Actor dependency.

Reliability and safety

• Stateful comparison between runs
• Duplicate-change protection
• Fetch-all-before-state-mutation fail-closed behavior
• Existing state preserved if any target fails
• Redirect destinations are revalidated
• Local/private/reserved network destinations are rejected
• Maximum 50 URLs per run
• 5 MB response-size limit per page
• Configurable normalized-text cap
• No login credentials collected
• No browser automation or paid AI API required

Limitations

V1 performs direct HTTP page fetching. Websites that render their useful content only through client-side JavaScript may require a future browser-enabled mode.

Some pages contain changing timestamps, counters, rotating content, or cookie banners. Use ignoreSelectors or contentSelector to reduce noisy changes.

Why use this instead of a raw page scraper?

A raw scraper answers:

"What does this page contain now?"

Website Change Intelligence Monitor answers:

"What changed since the previous check?"

That makes the output much more useful for ongoing competitor and business intelligence.

Responsible use

This Actor is intended for public pages. Users are responsible for ensuring their monitoring complies with applicable laws, website terms, and access policies.