How to Design a Counterfeit Risk Score for Brands

It is not possible to review thousands of digital listings with the same priority. A Counterfeit Risk Score is an analytical prioritisation model used to decide which records should receive expert assessment first.

The score is not a legal infringement decision. It helps the operations team direct its time towards high-risk cases.

Identify the risk factors

Price anomaly
Visual similarity
Seller history
Risky product description
Unauthorised channel
Past infringement
Entity-network connection

Weighting

Not every signal is equally important. Weights can be adjusted by examining previously verified cases for each brand and category.

Example risk tiers

An operational scale such as 0–100 can be used, with low-, medium- and high-risk groups, for example. Thresholds must be calibrated against real data.

Category calibration

Counterfeit signals differ between cosmetics, fashion and electronics. Category-specific coefficients may therefore be more appropriate than a single global model.

False-positive management

A genuine promotion can create a low-price anomaly, or an authorised seller may use catalogue images extensively. The model should therefore be continuously improved with expert feedback.

Human-review feedback loop

Cases confirmed or rejected by experts can be used as training data to improve model precision.

Model KPIs to track

Precision
False-positive rate
High-risk confirmation rate
Review time
Takedown success rate
Repeat infringement rate

Operationalisation and measurement note

When turning this topic into a real brand-protection operation, risk-factor identification, weighting and example risk tiers should be treated as parts of the same case record, not independent checks. For each detection, record the source, product, seller, channel, timestamp, risk rationale and action taken in structured fields wherever possible. This approach enables both individual case review and comparison of a seller’s or product group’s behaviour over time.

Instead of focusing on total alert volume, track the verified risk rate, false-positive rate, review time, repeat-seller rate and action outcomes together. Record the operational answer to “Does a Counterfeit Risk Score prove that a product is counterfeit?” and use expert feedback in subsequent detections. Over time, the monitoring system becomes decision support that helps the brand team prioritise cases more accurately, rather than merely finding more records.

Decision matrix and escalation

For sound decisions when designing a Counterfeit Risk Score for brands, insert a verification layer between detection and action. Risk-factor identification, weighting and risk tiers should be traceable on the same record. Even a strong signal must not produce a direct infringement conclusion without assessing business impact, data reliability and false-positive risk together. Low-confidence records can be directed to additional data collection, medium-confidence records to expert review, and high-confidence, high-impact records to the brand’s defined escalation process.

As operations mature, measure not only the number of records found but also which records become actual risks. Track source freshness, confirmation rate, false-positive rate, repeat sellers or accounts, average review time and action outcomes together. Linking a recurring risk to the previous case lets the team use earlier evidence and decisions instead of restarting its investigation. This turns isolated screenshots into an auditable, comparable risk-management system that can improve over time.

The B10 Digital Agency approach

B10 Digital Agency assesses counterfeit products, unauthorised sellers, brand impersonation and digital-channel risks alongside seller, product and channel data, rather than isolated listings. The aim is not just finding infringements but making recurring digital risks part of a measurable operational system.

Frequently asked questions

Does a Counterfeit Risk Score prove that a product is counterfeit?

No. The score only indicates operational risk and review priority.

What scale should the risk score use?

0–100 is a common, understandable example, but it is not mandatory.

Can every brand use the same model?

Generally not. Category and channel behaviour require calibration.

Can human review be removed?

Human verification must be retained for high-risk legal and commercial decisions.

Counterfeit Risk Scoring: An Analytical Model for Counterfeit Products and Unauthorised Sales
AI Counterfeit Detection: Combining Visual, Price and Seller Signals
Anti-Counterfeiting and Digital Brand Protection: Risk Mapping and Layered Defence

Measure your digital brand risk

Explore B10 Digital Agency’s brand-protection solutions to analyse marketplace, social-media and independent e-commerce risks at seller, listing and product level.

Explore Brand Protection & Monitoring →