Ad Pixel & Tech Stack Detector: Meta Pixel, GA4, GTM, Google Ads, Ad-Readiness Score

Paste a list of websites and get, for each one, the marketing and tracking tags it runs (Meta Pixel, GA4, Google Tag Manager, Google Ads conversion tracking, Microsoft UET, TikTok, LinkedIn, Pinterest, Snap, X, Reddit, Nextdoor, Hotjar, Clarity), the tag IDs that prove each one is really installed, its CMS or site builder, e-commerce and real-estate platform, call tracking, chat, booking, forms, CRM and email tools, the consent banner, and a 0–100 ad-readiness score with a list of gaps.

Built for:

• Ad agencies prospecting. "No pixel, no conversion tracking, no tag manager" is the pitch. Sort a list by score and call the bottom third.
• SEO and web agencies auditing a prospect's stack before a proposal.
• Lead-gen shops enriching Google Maps or directory lists of local businesses.
• AI agents that need a structured "what is this business running?" answer for one URL.

What you get

• Grouped answers: adPixels, analytics, tagManagers, callTracking, chat, booking, forms, crm, emailMarketing, consent, plus single best guesses for cms, siteBuilder, ecommercePlatform, realEstatePlatform.
• Tag IDs with evidence: Meta pixel IDs, GA4 measurement IDs, GTM container IDs, Google Ads conversion IDs and labels, UET, TikTok, LinkedIn, Hotjar, Clarity, HubSpot, CallRail and more. Every detection says how it was found (via), how sure we are (confidence) and shows a short evidence snippet.
• Google Tag Manager containers read for you. Most small-business tags are injected by GTM at runtime, so they are not in the HTML. The Actor fetches the site's public GTM container (the same file every visitor's browser loads) and recovers the Meta pixel, Google Ads conversions, GA4, UET, Hotjar, CallRail and custom-HTML tags inside it. Detections found this way are marked via: "gtmContainer".
• Platform settings read for you. Pixels added through Shopify's, Wix's or Squarespace's own marketing settings are read from the page (via: "platformConfig").
• Ad-readiness flags and score: noMetaPixel, noGa4, legacyUaOnly, noConversionTracking, noTagManager, noConsentBanner, noLeadCapture, noCallTracking, noLocalBusinessSchema, noGoogleBusinessProfileLink, slowSite, insecureSite, rolled into a score, a grade (A–F) and a gaps list ordered by points lost.
• Local-business signals: Schema.org LocalBusiness markup (type, city, region, ZIP and which fields are present), a Google Business Profile link, and which social networks the site links to.
• Site health: HTTPS and certificate, http-to-https redirect, HSTS, mixed content, time to first byte, page weight, third-party script hosts.

Sample output

One row per website (abbreviated; empty lists and most IDs trimmed):

{
  "inputUrl": "acme-plumbing.example",
  "finalUrl": "https://www.acme-plumbing.example/",
  "status": "ok",
  "scanMode": "http",
  "adReadinessScore": 95,
  "adReadinessGrade": "A",
  "cms": "WordPress",
  "wordpressBuilder": "Elementor",
  "adPixels": ["Meta Pixel", "Google Ads conversion tracking", "Microsoft Advertising UET"],
  "analytics": ["Google Analytics 4", "Hotjar"],
  "tagManagers": ["Google Tag Manager"],
  "callTracking": ["CallRail"],
  "forms": ["Contact Form 7"],
  "ids": {
    "metaPixelIds": ["1029384756473829"],
    "metaStandardEvents": ["Lead"],
    "ga4MeasurementIds": ["G-PTP4LMN8Q2"],
    "gtmContainerIds": ["GTM-PTP7Q2X"],
    "googleAdsConversionIds": ["AW-781234905"],
    "googleAdsConversionLabels": ["AW-781234905/Xy7_kLmNpQ"],
    "uetTagIds": ["187650432"],
    "callrailCompanyIds": ["456719823"]
  },
  "conversionSignals": [
    { "signal": "google-ads-conversion-label", "via": "gtmContainer" },
    { "signal": "meta-standard-event", "via": "gtmContainer" },
    { "signal": "call-tracking", "via": "gtmContainer" }
  ],
  "leadCapture": { "formTool": "Contact Form 7", "genericLeadForm": true, "bookingTool": null, "chatTool": null, "callTracking": "CallRail", "telLink": true, "mailtoLink": true },
  "schema": { "localBusinessFound": true, "types": ["Plumber"], "addressLocality": "Atlanta", "addressRegion": "GA", "postalCode": "30309", "hasTelephone": true },
  "googleBusinessProfile": { "linkFound": true, "linkType": "maps/place", "url": null },
  "site": { "https": true, "tlsOk": true, "httpRedirectsToHttps": true, "mixedContent": false, "ttfbMs": 412, "htmlBytes": 88213 },
  "adReadiness": {
    "score": 95,
    "grade": "A",
    "gaps": ["noConsentBanner"],
    "flags": { "noMetaPixel": false, "noConversionTracking": false, "noConsentBanner": true, "slowSite": false },
    "points": { "metaPixel": 20, "conversionTracking": 20, "ga4": 15, "tagManager": 10, "leadCapture": 10, "https": 10, "consent": 0, "speed": 5, "localPresence": 5 }
  },
  "containers": [{ "id": "GTM-PTP7Q2X", "kind": "gtm", "status": "inspected", "detections": 9 }],
  "technologies": [
    { "name": "Meta Pixel", "category": "adPixels", "confidence": "high", "via": "gtmContainer", "ids": ["1029384756473829"], "evidence": "connect.facebook.net/en_US/fbevents.js", "page": "/" },
    { "name": "Google Tag Manager", "category": "tagManagers", "confidence": "high", "via": "script", "ids": ["GTM-PTP7Q2X"], "evidence": "'GTM-PTP7Q2X'", "page": "/" }
  ]
}

Sites that block automated requests, are disallowed by robots.txt or fail are still listed (with status blocked, skipped or error and an errorCode) so you can see what happened, and they are not charged. A RUN_SUMMARY record in the run's key-value store counts sites per gap flag, the score histogram and the most common platforms.

Input

Field	What it does
Website URLs (urls)	Home pages or any public page, one per line. example.com works too. Up to 20,000 per run.
Extra pages (extraPaths)	Up to 3 paths scanned on every site as well, e.g. /contact, /book. Their detections are merged into the site's row.
Also scan the contact or booking page (autoFindContactPage)	Finds the first same-site link mentioning contact, book, schedule, appointment, quote or estimate and scans it.
Read Google Tag Manager containers (inspectTagContainers)	On by default. Recovers tags deployed through GTM. No extra charge.
Include evidence snippets (includeEvidence)	On by default.
Include the flat technologies list (includeTechnologiesList)	On by default; turn off for slimmer rows.
Respect robots.txt (respectRobotsTxt)	On by default. Turn off only for sites you own or may test.
Timeout, parallel sites, parallel requests per host	Advanced settings. Defaults: 15 s, 10 sites, 2 requests per host.

Pricing

Pay per event: you pay only for complete results.

Event	Free plan	Bronze	Silver	Gold
Site scanned (one website with a complete row)	$0.0035	$0.003	$0.0027	$0.0024
Extra page scanned (each extra path or contact page that returned 200 on the same site; pages that redirect elsewhere or back to a page already scanned are skipped and free)	$0.0012	$0.001	$0.0009	$0.0008
Site scanned in a browser (reserved for browser mode, coming in v1.1; not charged today)	$0.012	$0.01	$0.009	$0.008

That's from $3 per 1,000 sites, plus Apify's standard Actor start fee. Reading GTM containers is included. Not charged: sites that block us (403, 429, challenge pages), sites skipped by robots.txt, unreachable sites and errors. Set Max cost per run to cap any run; the Actor stops cleanly when it is reached.

Integrations and API

Run it from the Apify API, on a schedule, or from Make, Zapier, n8n and Google Sheets through Apify's integrations. Pipe the output of a Google Maps or directory scraper straight in, then sort by adReadinessScore. The Overview, Ad readiness, Tags & IDs and Platforms tabs export to CSV and Excel. AI agents can call it through Apify's MCP server: give it a URL, get back the stack, the IDs and the gaps.

Limits

• HTTP only (no browser) in this version. The Actor reads the HTML, response headers, the site's public Google Tag Manager containers and the marketing settings Shopify, Wix and Squarespace publish in the page. It does not run the page's JavaScript. It can therefore miss tags that are added by other scripts at runtime (a theme's own tag loader, Segment or Zaraz destinations, custom code that builds IDs on the fly), tags that only load after a cookie-consent click, server-side tagging behind the site's own domain, and tags on single-page apps whose HTML is nearly empty. Such pages get renderHint: "client-rendered"; every row shows scanMode.
• Coverage has not been measured yet. Reading GTM containers and platform settings finds many tags the HTML alone misses, but some tags are only visible to a real browser; browser mode (v1.1) is designed to close that gap.
• "No pixel found" means none was found by these methods, not proof that none exists. Check containers and renderHint before pitching.
• Only published GTM tags are visible (paused tags are not, which avoids false positives). Custom-HTML tags that assemble IDs from variables at runtime are not.
• Focused catalog: about 120 signatures covering the tools agencies care about, not thousands of generic technologies.
• Sites behind bot protection are reported as blocked and never retried. No proxies are used.
• The ad-readiness score is a heuristic for prioritising outreach, not an audit. Each row's adReadiness.points shows the points earned per component, and they add up to the score.

FAQ

How is the score calculated? Points are awarded, not deducted, so a blank site scores near zero: Meta Pixel 20, conversion tracking 20, GA4 15 (Universal Analytics only: 5), tag manager 10, a lead-capture path 10, clean HTTPS 10 (valid certificate 6, http redirects to https 2, no mixed content 2), consent banner or Consent Mode 5, speed 5 (first byte within 1 s: 3; page under 1.5 MB with fewer than 20 third-party script hosts: 2), LocalBusiness schema or a Google Business Profile link 5. Grades: A 85+, B 70–84, C 50–69, D 30–49, F under 30. The flags alone can't reproduce the score (for example, insecureSite is one flag but HTTPS points come in parts), so use adReadiness.points: the points earned per component, which add up to score. scoreVersion changes if the weights ever do.

What counts as conversion tracking? A Google Ads conversion ID or label, a hard-coded Meta standard event (Lead, Purchase, Contact, Schedule and similar), Microsoft UET, the TikTok pixel, or a call-tracking tool. A label (AW-…/…) means a conversion action is configured; an AW- ID alone is reported with medium confidence.

What does via mean? script and html: hard-coded in the page. header and cookie: from the response. gtmContainer / gtagContainer: installed through Google Tag Manager or a Google tag. platformConfig: added in Shopify, Wix or Squarespace settings.

Why did my own site come back blocked? Your firewall or CDN refused our request. Allowlist the SaiberOpsBot User-Agent and run it again; blocked rows cost nothing.

Why are some fields booleans instead of values? Some page details can identify a person (a Calendly link is often someone's name, a Maps link can carry an owner's name, phone numbers and email addresses belong to people). Those are reported as present or absent only.

Is a sole trader's pixel ID personal data? Tag IDs identify an advertising or software account, which is a business identifier, and they are served to every visitor of the site. The Actor never builds a reverse lookup ("which other sites use this pixel").

Responsible use and data sources

The Actor reads only public pages at URLs you supply (the input page, up to three extra paths you name, optionally one contact page), each host's robots.txt, and the public Google Tag Manager container files those pages already tell every browser to load. It identifies itself as SaiberOpsBot, respects robots.txt by default (RFC 9309), never logs in, never accepts cookies, never solves CAPTCHAs, and does not rotate IP addresses. It refuses private, loopback and cloud-metadata addresses, including after redirects.

Output is limited to facts about the business's website: technology names, tag and account IDs, platform names, presence flags, city/region/ZIP from the site's own LocalBusiness markup, hostnames and HTTP metadata. It never outputs people's names, email addresses, phone numbers, street addresses, Calendly slugs, Maps place names, page text, reCAPTCHA site keys or account tokens; a final check redacts anything shaped like an email address or phone number before a row is saved. The detection catalog is original work written from vendors' public installation instructions.

Support

Found a bug or need a technology added? Open an issue on this Actor's Issues tab and we'll respond there.

Maintained by Saiber Operations LLC.