Google AI Overview & Brand Visibility Tracker (GEO)

Track how visible your brand is in Google's AI Overview (the generative "AI Overview" answer box, formerly SGE) and in classic organic results - for any list of search queries. Built for GEO (Generative Engine Optimization) and SEO monitoring: know when an AI Overview appears, which sources it cites, and whether your domain is one of them.

What you get (one record per query)

• ai_overview_present - is there an AI Overview on the page?
• ai_overview_text - the AI Overview's answer text (cleaned of citation badges)
• ai_overview_sources - the list of domains the AI Overview cites (the core GEO signal), plus ai_overview_source_urls (the full cited URLs)
• tracked_domain_present_in_overview - is your domain cited inside the AI Overview? (when trackDomain is set)
• organic_results - top 10 organic results: position, title, url, domain
• tracked_domain_organic_rank - your domain's organic position (or null)
• people_also_ask - the "People also ask" questions
• serp_state - results on success (or js_challenge / consent / captcha if a proxy misconfiguration got a gate instead of results)

Input

• queries (required) - the searches to track, e.g. ["best crm for small business", "how to file an llc"]. One SERP is fetched per query.
• trackDomain - your website domain, e.g. acme.com. Enables the "cited in AI Overview?" and organic-rank fields (subdomains match too).
• trackBrand - your brand name (stored on each record for reporting).
• gl / hl - country and language codes for localisation (default us / en).
• maxItems - cap on how many queries to process.
• proxyConfiguration - keep the default GOOGLE_SERP proxy group (see below).

Example input

{
  "queries": ["best crm for small business", "generative engine optimization"],
  "trackBrand": "Zoho",
  "trackDomain": "zoho.com",
  "gl": "us",
  "hl": "en",
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["GOOGLE_SERP"] }
}

Example output (one query)

{
  "query": "best crm for small business",
  "timestamp": "2026-07-10T14:40:00Z",
  "ai_overview_present": true,
  "ai_overview_text": "The best CRM for a small business depends on budget and needs...",
  "ai_overview_sources": ["hubspot.com", "zoho.com", "forbes.com"],
  "tracked_domain_present_in_overview": true,
  "organic_results": [
    { "position": 1, "title": "HubSpot CRM...", "url": "https://www.hubspot.com/products/crm", "domain": "hubspot.com" },
    { "position": 3, "title": "Zoho CRM...", "url": "https://www.zoho.com/crm/", "domain": "zoho.com" }
  ],
  "tracked_domain_organic_rank": 3,
  "people_also_ask": ["Which CRM is best for a small business?"],
  "serp_state": "results"
}

Why this Actor

• AI Overview-first. Most SERP scrapers only give you organic results. This one is built around detecting the AI Overview and extracting the domains it cites - the metric GEO/SEO teams now care about most.
• Durable parser. Google rotates its CSS class names every few weeks, so detection keys on stable signals (the visible "AI Overview" label, data-attrid, and the <a><h3> organic pattern) with class-name hints only as a fallback.
• Honest reporting. Every record carries serp_state, so a proxy/consent gate is never silently reported as "no AI Overview".

Proxy & how Google fetching works

Google serves a JavaScript/bot gate to raw IPs, so this Actor uses Apify's GOOGLE_SERP proxy group (available on the free plan, tuned for Google SERP scraping) to get clean results HTML. Keep the default proxy configuration. Requests go to http://www.google.com/search (the GOOGLE_SERP proxy is HTTP-only and requires the www. host); localisation is via gl/hl.

AI Overview caveat (important): an AI Overview can be (1) rendered in the initial HTML, (2) deferred and injected by JavaScript after the page paints, or (3) absent. This Actor fetches HTML over HTTP, so it reliably catches states (1) and (3). A minority of overviews load only via JavaScript (state 2) and will read as ai_overview_present: false - this is a known limitation of every non-headless SERP scraper. Organic results, source detection, and rankings are unaffected.

This data is for market/SEO research. It is not a consumer report and must not be used for FCRA-covered purposes.