Schema SEO Regression Monitor

Monitor structured data changes across your website and detect SEO-critical schema regressions. Know when Product, Offer, AggregateRating, Article, or Breadcrumb markup disappears — before it hurts your search visibility.

The Problem This Solves

You deployed a website update. Did it remove important structured data?

Generic schema validators tell you what schema exists right now. Generic change detectors tell you something changed. Neither tells you what matters:

• Did a Product lose its Offer markup?
• Did AggregateRating disappear?
• Did an Article lose its author?
• Is this change a false positive (just key reordering)?

This Actor answers: WHAT changed, WHY it matters, HOW important it is, and WHAT to investigate first.

Who Is It For?

• SEO agencies — Monitor client websites for schema regressions after deployments
• Website owners — Catch structured data issues before they affect search appearance
• Developers — Verify deployments don't accidentally remove schema markup
• Automated pipelines — Schedule runs via Apify tasks for continuous monitoring

How It Works

Website URL
     ↓
Fetch HTML (HTTP, no browser)
     ↓
Extract JSON-LD + Microdata + RDFa
     ↓
Normalize (semantic, not textual)
     ↓
Load previous snapshot from KVS
     ↓
Compare (semantic diff)
     ↓
Classify changes by SEO impact
     ↓
Generate recommendations
     ↓
Push structured results to dataset

Key Design Principles

1. Semantic comparison — Key order changes don't trigger false positives
2. Value preservation — Price changes (100→120) are always detected
3. Snapshot persistence — Same URL = same snapshot key across runs
4. Failure safety — Fetch failures preserve previous snapshot (never false "schema removed")
5. Baseline first — First run creates baseline, no false regressions

Supported Schema Types

The Actor recognizes SEO importance for:

Schema Type	Importance	Removal Impact
Product	CRITICAL	HIGH/CRITICAL
Offer	CRITICAL	HIGH/CRITICAL
AggregateRating	CRITICAL	HIGH/CRITICAL
Review	HIGH	HIGH
FAQPage	HIGH	HIGH
Article / NewsArticle	HIGH	HIGH
BreadcrumbList	MEDIUM	MEDIUM
LocalBusiness	MEDIUM	MEDIUM
Event	MEDIUM	MEDIUM
JobPosting	MEDIUM	MEDIUM
Recipe	MEDIUM	MEDIUM
VideoObject	MEDIUM	MEDIUM
Organization	MEDIUM	MEDIUM

Input

{
  "urls": ["https://example.com/product/1", "https://example.com/product/2"],
  "snapshotNamespace": "client-acme",
  "maxUrls": 100,
  "requestDelayMs": 500,
  "enableMicrodata": true,
  "enableRdfa": true
}

Or use a sitemap:

{
  "sitemapUrl": "https://example.com/sitemap.xml",
  "snapshotNamespace": "client-acme",
  "maxUrls": 500
}

Output

Each URL produces one dataset item:

{
  "url": "https://example.com/product/iphone",
  "status": "CHANGE_DETECTED",
  "previousSchemaTypes": ["Product", "AggregateRating", "BreadcrumbList"],
  "currentSchemaTypes": ["Product", "BreadcrumbList"],
  "changeCount": 2,
  "changes": [
    {
      "type": "SCHEMA_TYPE_REMOVED",
      "schemaType": "AggregateRating",
      "impact": "CRITICAL",
      "priority": "HIGH",
      "confidence": "HIGH",
      "changeScore": 80.0,
      "recommendation": "Verify whether AggregateRating schema was intentionally removed..."
    }
  ],
  "highestPriority": "HIGH",
  "priorityScore": 80.0,
  "fetchStatus": 200,
  "timestamp": "2026-08-30T18:00:00Z",
  "snapshotId": "abc123",
  "previousSnapshotPreserved": false
}

Status Values

Status	Meaning
BASELINE_CREATED	First run — snapshot saved, no comparison
NO_CHANGE	Schemas match previous snapshot
CHANGE_DETECTED	Semantic differences found
FETCH_FAILED	Page returned same error as before — previous snapshot preserved
FETCH_STATUS_CHANGED	Page status changed (e.g. 200→500) — previous snapshot preserved

Priority Methodology

Change Score = Schema_Importance × Change_Severity × Property_Factor

Score Range	Priority	Meaning
75-100	CRITICAL	Immediate investigation needed
50-74	HIGH	Likely affects structured data availability
25-49	MEDIUM	Worth investigating
0-24	LOW	Minor or positive change

Priority scores are SEO regression prioritization heuristics, not Google ranking scores.

Limitations

• The Actor does not guarantee Google rich-result eligibility
• The Actor does not guarantee ranking changes
• Structured-data interpretation may change over time
• Schema.org support and search-engine feature requirements are not identical
• The Actor detects changes; it does not determine whether a change was intentional

Pricing

Pay-per-event: $0.10 per schema-regression-analysis event

Charged only for successfully analyzed URLs. Failed fetches and invalid URLs are not charged.

License

MIT