/ writing / automation

Diff-based page monitoring: the boring architecture that just works

2026-07-27 · cynix · 3 min read

Most page-monitoring tools try to be smart: semantic diffing, DOM comparison, visual regression. The problem is "smart" breaks silently when the site redesigns. The approach that survives redesigns is dumb: fetch, store, diff lines, alert on change.

Page Change Monitor does exactly that. Point it at a URL, schedule it, and it emits structured change records:

{
  "url": "https://example.com/pricing",
  "checkedAt": "2026-07-27T08:00:00Z",
  "changed": true,
  "changeType": "modified",
  "diff": {
    "added": ["  $49/mo"],
    "removed": ["  $39/mo"]
  },
  "snapshotHash": "sha256:abc123..."
}

Why line diff beats semantic diff

A CSS class rename breaks DOM diffing. A wrapper

breaks tree diffing. A line diff catches the actual text change regardless of structure. You get false positives when whitespace shifts, but you never get false negatives — the thing you actually care about (did the price change?) always surfaces.

What the actor handles

Full-page snapshots with configurable selectors (watch a price element, ignore the nav bar)
Automatic hash comparison — no storage bloat on unchanged pages
Runs on Apify's schedule (cron-style, down to hourly)
Webhook, email, or dataset push for alerts
Dedup by URL + hash — re-runs don't spam you

It's not clever. It's reliable. That's the product.

Try it: Page Change Monitor on Apify — free tier covers most personal use cases.