Home/Offers/E-commerce & Retail/Counterfeit Listing Detection API

E-commerce & Retail

Counterfeit Listing Detection API

Counterfeit sellers are not subtle at scale. They reuse the same images, cluster around the same price points and recycle the same seller patterns — which is precisely what makes them detectable.

Marketplaces
Brand protection teams
Trust and safety teams
Commerce platforms
Brand manufacturers

Discuss This Build See What It Returns

What your customers would see

example interface

MockupIllustrative mockup of the interface we would build. Figures shown are examples, not results from a live deployment.

The problem this solves

A marketplace with open seller onboarding acquires a counterfeit problem the moment it becomes worth exploiting. The enforcement response is almost always reactive: a brand files a complaint, a trust-and-safety analyst reviews the listing, it comes down, and the same seller relists within the day under a new account with the same photographs.

The detectable patterns are consistent. Product imagery lifted from the brand's own site or from other listings. Prices well below any plausible authorised cost. Titles stuffed with brand terms for products the seller has no business carrying. New accounts whose entire catalogue appeared in one upload. Shipping origins that do not match the claimed stock location.

None of those signals is proof, and that distinction is the whole design constraint. A counterfeit accusation against a seller has legal and commercial consequences, and grey-market goods, authorised resellers and parallel imports all look superficially similar to counterfeits. The output has to be a prioritised review queue with evidence attached, not an automated takedown.

Who buys this

Marketplaces, Brand protection teams, Trust and safety teams, Commerce platforms, Brand manufacturers

Category: E-commerce & Retail

What the API returns

What comes back

Risk score per listing with the contributing signals itemised
Image reuse detection: perceptual matches against brand assets and across other listings
Price anomaly detection against the authorised and observed market range
Brand-term misuse in titles, descriptions and attributes
Seller-level risk aggregation, including catalogue upload patterns and account age
Listing-cluster detection, which links relisted inventory across accounts
Evidence bundles formatted for a brand notice or an internal enforcement record
Explicit grey-market and authorised-reseller flags, to separate them from counterfeit suspicion

Endpoints

Method	Path	Returns
POST	/v1/listings/score	Score one or many listings for counterfeit risk
POST	/v1/images/match	Perceptual image match against brand assets and known listings
GET	/v1/sellers/{id}/risk	Aggregated seller-level risk with contributing listings
GET	/v1/queue	Prioritised review queue with evidence bundles
POST	/v1/brands/{id}/assets	Register authorised imagery, price ranges and reseller lists
POST	/v1/outcomes	Record enforcement decisions to recalibrate the queue

Request and response

illustrative

Request

{
  "listing": {
    "id": "lst_99120344",
    "title": "Genuine Oakwell Leather Holdall 100% Authentic UK Stock Fast Ship",
    "brand_claimed": "Oakwell",
    "price": 74.00,
    "currency": "GBP",
    "images": ["https://cdn.example.com/l/99120344_1.jpg"],
    "seller_id": "slr_41288",
    "ships_from": "CN"
  },
  "include": ["image_matches", "evidence"]
}

Response

{
  "listing_id": "lst_99120344",
  "risk_score": 88,
  "band": "high",
  "recommended_action": "queue_for_human_review",
  "signals": [
    { "signal": "image_matches_brand_asset", "detail": "Perceptual match to Oakwell product photography", "weight": 0.31 },
    { "signal": "price_below_authorised_floor", "detail": "74.00 against observed floor of 219.00", "weight": 0.27 },
    { "signal": "authenticity_keyword_stuffing", "detail": "'genuine', '100% authentic' in title", "weight": 0.14 },
    { "signal": "ship_origin_mismatch", "detail": "Claims UK stock, ships from CN", "weight": 0.13 },
    { "signal": "seller_bulk_upload", "detail": "412 listings created within 90 minutes", "weight": 0.09 }
  ],
  "image_matches": [
    { "source": "brand_asset_library", "similarity": 0.94, "asset_id": "oak_holdall_03" },
    { "source": "removed_listing", "similarity": 0.91, "listing_id": "lst_88104221" }
  ],
  "alternative_explanations": [
    "Parallel import of genuine stock",
    "Unauthorised but genuine reseller clearing old season"
  ],
  "seller_risk": { "score": 81, "listings_flagged": 37, "account_age_days": 14 }
}

Where the data comes from

Your listing data	Titles, descriptions, attributes, images, prices and seller identifiers from your own marketplace
Brand-supplied asset libraries	Authorised product photography, authorised price ranges and reseller lists, provided by the brands under a cooperation agreement
Seller account and behavioural data	Account age, upload patterns, address and payout signals — held by you, and among the strongest predictors
Enforcement history	Previously removed listings and confirmed counterfeits, which are the training labels and the image corpus for relisting detection
Trademark and brand registries	Official registrations, which establish which brand terms are protected and by whom
Permitted external listing data	Where cross-marketplace comparison is wanted, only from sources whose terms allow collection and use for this purpose

On data licensing. Being allowed to read a dataset does not automatically mean you may redistribute it, resell it through an API, or train a model on it. Before any build starts we check the licence, terms and derived-data rules of every source, and we design around the ones that do not permit your intended use. Where a commercial licence is required, that cost is identified during discovery rather than discovered later.

The intelligence layer

The difference between a demo and a service is mostly in this section. These are the parts that keep the answer trustworthy once real traffic is hitting it.

Perceptual image hashing for near-duplicate detection, plus learned embeddings for the cases where an image has been cropped, mirrored or watermarked out
Price anomaly detection against a per-product authorised and observed distribution rather than a flat percentage rule
Brand-term misuse detection that distinguishes compatible-with and replacement-for phrasing from false authenticity claims
Seller-level graph features linking accounts by image reuse, payout details and upload timing, which is how relisting networks become visible
Explicit modelling of benign explanations — parallel imports, authorised clearance, legitimate used goods — so they are surfaced alongside the suspicion
Calibration and queue ranking by expected enforcement value, so analyst time goes where it recovers most
Precision measured against analyst decisions, with false-positive cost tracked separately from false-negative cost
Monitoring for evasion drift, because sellers adapt to whatever signal was last enforced on

What it changes for the business

Analyst throughput

A ranked queue with evidence attached means reviewers decide rather than investigate from scratch.

Faster relisting shutdown

Image and account linkage catches the same inventory reappearing under a new seller within hours.

Brand relationships

Proactive detection and structured evidence bundles change the conversation with rights holders.

Defensible decisions

Every enforcement action carries a recorded signal trail, which matters if a seller disputes it.

How it fits into your product

How it embeds

Synchronous scoring at listing creation, so high-risk items can be held before they go live
Batch scoring across the existing catalogue for backlog clearance
Queue API your existing trust-and-safety tooling can render, with evidence bundles as structured JSON
Brand portal endpoints so rights holders can register assets and price floors themselves
Outcome feedback endpoint that recalibrates ranking from your analysts' actual decisions
Deployment inside your environment, which is normal here given the sensitivity of seller data

Included in every build

Authentication, scoped API keys and per-customer rate limiting
Input validation with predictable, documented error responses
Versioned endpoints so your integration never breaks without notice
Request and response logging for audit and dispute resolution
Latency, error-rate and usage monitoring with alerting
OpenAPI specification, sandbox environment and client examples
Backups, failover design and a documented recovery procedure
An agreed support arrangement once the service is live

How we would build it

about 20 weeks

Discovery and feasibility

Weeks 1–3

This one depends heavily on data access, so we start by proving the sources exist, can be licensed for your use, and actually carry the signal the model needs. If they do not, you find out here.

Architecture and baseline model

Weeks 4–6

A written architecture and a baseline model measured against a holdout period, so you know what accuracy you are buying before the full build starts.

Production build

Weeks 7–16

Data pipelines, model training and serving, historical archive, validation harness, API, security, monitoring and documentation.

Validation and integration

Weeks 17–20

Backtesting against held-out history, accuracy reporting your team can audit, integration support and an agreed retraining cadence.

Commercial models

There is no fixed price list, because the cost is driven by data complexity, integration count, model requirements, request volume and the support level you need. Scope is agreed in writing before development starts.

Fixed-scope prototype

A defined slice of the service, built and priced as one piece of work. The usual first step when the business case needs proving internally.

Fixed-price build

An agreed scope against a written architecture, delivered for an agreed price. Best when the requirements are already clear.

Build then retainer

We build the service, then operate and improve it under a monthly agreement covering monitoring, updates and model maintenance.

White-label product

The service runs under your brand, your domain and your pricing, with MLPivot invisible to your customers.

Questions buyers ask

Can it automatically remove listings?

We would strongly advise against it and would not design it as the default. Counterfeit determination has legal weight, and genuine parallel imports, authorised clearance and legitimate second-hand goods all trip the same surface signals. The API ranks and evidences; a human decides. Automated holds at listing creation for the very highest band are defensible, full automated takedown is not.

What are the false-positive risks?

Real and asymmetric. Wrongly removing a legitimate seller's listing costs you a seller relationship and may cost you a legal argument. That is why the response includes alternative explanations and why we track false-positive cost separately rather than optimising a single accuracy number.

Do you need the brands to cooperate?

It helps enormously. Authorised imagery, authorised price floors and reseller lists turn weak signals into strong ones. The system works without them on behavioural and image-reuse signals alone, but detection quality on a given brand roughly tracks how much that brand has supplied.

Can you scan other marketplaces for our brand?

Only where the source permits it. Cross-marketplace monitoring is frequently requested and frequently constrained by terms of use, and some platforms offer an official brand-protection programme that is the correct route instead. We establish per source during discovery what may be collected and how the results may be used.

Will counterfeiters adapt to this?

Yes. They adapt to every detection system, usually within weeks of enforcement becoming consistent. That is why the build includes evasion monitoring and a retraining cadence rather than a one-off model, and why the signal mix is deliberately broad instead of resting on image matching alone.

All E-commerce & Retail offers →

Want this built for your platform?

Tell us what your product does today and what you want another system to be able to ask it. We will come back with an approach, a scope and a price.

Discuss This Build Browse All Offers