Brand Protection: Counterfeit, Fake Shop & Typosquat Detector

What does Brand Abuse & Counterfeit Detector do?

Brand Abuse & Counterfeit Detector finds the websites, listings, domains and profiles that are trading on your brand name and tells you which ones are worth a takedown. Give it your brand and your official domains, and one run sweeps search results, marketplaces, social platforms and lookalike domain space, then uses AI to classify every hit as a fake webshop, a counterfeit listing, a typosquatted domain, an impersonation profile, a legitimate retailer or plain noise. Confirmed infringements come back with a risk score, the evidence behind the verdict, a dated HTML snapshot of the page and the takedown route to use.

It runs on Apify, so you get a scheduler for weekly sweeps, an API and webhooks to push findings into your own systems, proxy rotation so the sites you are investigating do not block you, and run monitoring with alerts. No API key is required: the AI access is built in.

Why use Brand Abuse & Counterfeit Detector?

Counterfeiting is a trillion-dollar problem and the tooling around it is priced for legal departments, not for the brands actually losing the sales. Commercial brand protection platforms start in the tens of thousands per year and lock the data behind their own dashboards. This Actor does the discovery and triage part for cents per candidate, and hands you a plain dataset.

Use it to:

• Find fake webshops built to look like your outlet store, before your customers get burned and blame you.
• Catch counterfeit marketplace listings on AliExpress, DHgate, eBay, Etsy, Temu and the rest, with the goods and seller page captured.
• Monitor typosquatted domains: misspellings, character swaps, hyphenations, added words and alternative TLDs of your own domain, checked for live DNS.
• Spot impersonation profiles posing as your official store on Facebook, Instagram, TikTok, X, Telegram and YouTube.
• Build a takedown queue with dated evidence, so registrar complaints, UDRP filings, marketplace notices and platform reports can be filed without re-doing the research.
• Prove the trend over time: schedule a weekly sweep and watch whether enforcement is actually reducing the count.

Typical buyers: DTC and ecommerce brands, IP counsel and trademark agencies, marketplace sellers defending a registered brand, and agencies running brand protection as a service.

How to use Brand Abuse & Counterfeit Detector

1. Click Try for free and open the Input tab.
2. Type your brand name exactly as shoppers search for it.
3. Add every domain you own in Official domains. This keeps your own sites out of the results and gives the lookalike detector its baseline.
4. Pick which checks to run. Fake shops and marketplaces are the highest-yield starting point.
5. Click Start. Watch the run log, or come back to the Output tab when it finishes.
6. Export the results as JSON, CSV, Excel or HTML, or pull them from the API.
7. Open Schedules and set it to run weekly. Brand abuse is a recurring problem, not a one-time audit.

Input

Only the brand name is required. Everything else has a working default. The full list is on the Input tab; the fields that matter most:

Field	What it does
brandName	The trademark to sweep for.
officialDomains	Domains you own. Excluded from results and used as the typosquat baseline.
checks	Which abuse types to look for: fake shops, marketplace listings, lookalike domains, social impersonation.
productKeywords	Product lines counterfeiters copy most, for extra targeted searches.
brandContext	Free-text notes so the AI can judge borderline cases the way you would.
maxCandidates	Hard cap on how many candidates get screened. Your main cost control.
minRiskScore	The AI risk score at which a hit counts as a confirmed infringement.
countryCode	Which country the searches run from. Sweep each key market separately.

{
  "brandName": "Ray-Ban",
  "officialDomains": ["ray-ban.com", "rayban.com"],
  "checks": ["fake_shops", "marketplace_listings", "typosquat_domains"],
  "productKeywords": ["Wayfarer", "Aviator"],
  "maxCandidates": 120,
  "minRiskScore": 60,
  "countryCode": "us"
}

Output

One row per screened candidate. Download as JSON, HTML, CSV or Excel, or read it straight from the Apify API.

{
  "brand": "Ray-Ban",
  "url": "https://ray-bn.com/",
  "domain": "ray-bn.com",
  "registrableDomain": "ray-bn.com",
  "discoveredVia": "typosquat",
  "verdict": "typosquat_domain",
  "verdictLabel": "Typosquatted domain",
  "isInfringement": true,
  "riskScore": 90,
  "confidence": "high",
  "reasoning": "A one-character deletion from the official domain, registered and live, redirecting traffic to an unrelated commercial site.",
  "evidenceSignals": ["one-character-deletion", "redirects-off-domain", "no-brand-affiliation"],
  "counterfeitGoods": [],
  "recommendedAction": "File a registrar abuse report and prepare a UDRP complaint for the domain.",
  "ipAddress": "34.111.46.214",
  "httpStatus": 200,
  "evidenceUrl": "https://api.apify.com/v2/key-value-stores/.../records/evidence-ray-bn.com-8c634c7165.html",
  "checkedAt": "2026-07-29T14:07:02+00:00"
}

Data fields

Field	Description
url	The page or domain that was screened.
domain / registrableDomain	Hostname, and the domain as a registrar complaint would name it.
discoveredVia	Which sweep surfaced it: fake_shop, marketplace, social or typosquat.
verdict / verdictLabel	The AI classification and its plain-English label.
isInfringement	True when the verdict is brand abuse and the risk score cleared your threshold.
riskScore	0 to 100 confidence that this harms the brand.
confidence	How sure the AI is of its own verdict.
reasoning	Short explanation citing what was actually on the page.
evidenceSignals	The specific tells behind the verdict.
pageSignals	Automated tells found on the page: official claims, deep discounts, missing legal pages, wire-transfer-only payment.
counterfeitGoods	Products or categories offered under your brand name.
recommendedAction	The takedown route a brand protection team would use.
evidenceUrl	Dated HTML snapshot of the page, for confirmed infringements.
ipAddress	Address a lookalike domain resolves to.
httpStatus	Status returned when the page was opened.
checkedAt	UTC timestamp your evidence is stamped with.

How much does brand protection monitoring cost?

This Actor charges per event: a small fee for each candidate it screens, and a larger one for each infringement it confirms with evidence. Discovery, DNS checks and de-duplication are free, so a sweep that generates 400 lookalike domains and finds 9 registered ones only ever bills for those 9.

maxCandidates is the ceiling on a run's cost. Set it to 120 for a normal weekly sweep, higher for a first full audit of a heavily counterfeited brand.

AI usage is billed separately by Apify's model integration, not by this Actor, so you are never charged twice for the same tokens. The price per million tokens is shown next to every model in the dropdown. If you would rather pay OpenRouter directly, drop your own OpenRouter key into the input and the calls bypass Apify billing entirely.

Free Plan limits: free-plan runs screen at most 10 candidates from 4 searches, read 1 page of results per search and generate 30 lookalike domains. That is enough to see the output shape, not enough to protect a brand. Subscribe to an Apify plan to lift the caps. Note that the Free Plan is also charged roughly 10x by the model integration for identical AI usage, so a paid plan cuts that cost too.

Tips

• Sweep each market separately. Counterfeit storefronts are targeted by region. Run one schedule per countryCode for your key markets rather than one global sweep.
• List every domain you own, including the misspellings you defensively registered. Anything you leave out will come back as a typosquat finding.
• Use brandContext to teach it your market. Naming your authorized retailers and your real price range cuts false positives more than any other setting.
• Turn on onlyInfringements once you trust the output, so the dataset is a takedown queue rather than a full audit trail.
• Lower minRiskScore for the first run to see the grey area, then raise it for the recurring schedule.
• Leave residential proxy enabled. Fake shops routinely block datacenter traffic, and a blocked fetch means the AI judges a snippet instead of the page.

FAQ

Is monitoring for counterfeits and brand abuse legal? Yes. This Actor reads publicly available pages, the same ones any shopper can open. It collects no personal data and does not log in anywhere. Whether a specific page infringes your rights is a legal question for you and your counsel; this Actor gathers the evidence, it does not give legal advice.

Does the AI verdict replace a lawyer's review? No. It is a triage layer that turns thousands of search results into a short, ranked list a human can actually review. Always check the reasoning and the snapshot before filing.

Do I need an OpenAI or OpenRouter API key? No. AI access is built in and needs no key from you. Supplying your own OpenRouter key is optional and only changes who gets billed for the tokens.

How do I find fake shops using my brand name? Enable the fake shops check, add your product lines to productKeywords, and run a sweep per market. The Actor searches the phrasings counterfeiters actually use (outlet, clearance, official store, deep discounts) while excluding your own domains.

How do I check for typosquatted domains? Add your domains to officialDomains and enable the lookalike domains check. The Actor generates misspellings, doubled and dropped letters, transpositions, keyboard-neighbour typos, homoglyphs, hyphenations, added words like "outlet" or "shop", and alternative TLDs, then keeps only the ones with live DNS.

Can I run this from an AI agent? Yes. Call it from any agent through the Apify MCP server, or from the Apify API. The output schema is published, so an agent can read and chain the results without extra glue.

Can it file the takedowns for me? No. It produces the evidence pack and names the route (registrar abuse report, UDRP, marketplace notice, platform impersonation report). Filing stays with you or your counsel.

Why did a legitimate retailer show up in my results? Because it sells your goods and the sweep found it. It is classified as authorized_retailer or unauthorized_reseller, not as infringement. Add it to ignoredDomains to stop screening it on future runs.

The run found nothing. What now? Widen checks, add productKeywords, and raise maxCandidates and pagesPerQuery. If the log shows Google refusing the searches, the run could not reach Apify's Google SERP proxy: that needs Apify Proxy access on your plan, and without it the sweep falls back to a backup engine that returns considerably less.

Which search engine does it use? Google, through Apify's dedicated Google SERP proxy, so the results are the ones your customers actually see. A backup engine sits behind it purely as a safety net for runs that cannot reach the SERP proxy. The run summary reports searchesBlocked and searchesViaBackupEngine so you always know which one produced your findings. The proxy setting in the input governs how candidate pages are opened and does not affect searching.

Support

Found a bug or a false positive pattern worth fixing? Open an issue on the Issues tab and it gets looked at.

⚡ Unlock full pages and higher limits by subscribing to an Apify plan. 💼 Need a custom solution? Reach out at youssefarhan24@gmail.com 🌐 Got a site in mind? Suggest it here

Disclaimer

This Actor collects publicly accessible information for brand protection and anti-counterfeiting purposes. It does not access private data, bypass authentication, or interact with the sites it inspects beyond loading a public page. AI classifications are automated judgments and can be wrong: review the reasoning and the saved snapshot before acting on any finding. Nothing here is legal advice.