Soft Brand Mismatch

Soft Brand Mismatch occurs when a merchant's public branding is technically accurate but semantically misleading, effectively lowering their perceived risk profile while targeting a high-risk customer segment.

---

📝 Short Summary

• Scenario: A shop selling "Android Streaming Sticks" (Low Risk) that are actually pre-loaded with unauthorized IPTV apps (High Risk).
• Business Motivation: To pass automated underwriting by appearing as a standard electronics retailer (MCC 5732) rather than a digital content subscription service.
• Key Deception: The hardware is the "Trojan Horse" for the high-risk software/service.

---

🏗 Technical Architecture

There is no technical cloaking (no redirects, no hidden pages). The deception is in the product catalog curation.

Frontend Behavior (Customer View)

1. Landing Page: Professional e-commerce site selling "4K Media Players".
2. Product Page: "SuperStream Box - Unlocked - Access all your favorite media."
3. Reviews: Users comment "Great for watching sports for free!" (Red Flag).

Backend Behavior (PSP View)

• MCC: 5732 (Electronics Stores).
• Ticket Size: $50 - $100 (Hardware cost).
• Line Items: "Model X Media Player".

---

🕵️‍♂️ Detection Challenges

• Legitimate Appearance: The merchant is shipping a physical product. Proof of Delivery (tracking numbers) will be valid.
• Keyword Evasion: They avoid words like "IPTV", "Channels", "Subscription". They use "Unlocked", "Jailbroken", "Loaded".
• Automated Pass: Most crawlers see valid hardware listings and approve the account.

---

🏦 PSP Detection Probability

Provider	Probability	Detection Analysis
Stripe	75%	Medium/Strong. Radar detects semantic clusters in product descriptions (e.g., "unlocked" + "stick").
Amazon Pay	90%	Very Strong. Amazon knows this product category intimately and aggressively bans "jailbroken" devices.
PayPal	65%	Medium. Often detected post-transaction via disputes mentioning "channels not working".
Shopify Payments	85%	Strong. AUP (Acceptable Use Policy) bots scan for terms like "kodi loaded" or "unlimited movies".
Square	60%	Medium. Retail-focused; may miss the nuance of the software inside the hardware.
Authorize.net	40%	Weak. Relies on the ISO/Acquirer to spot the inventory nuance.

---

🛡️ Recommended Detection Strategies

1. Sentiment & Keyword Analysis on Reviews

Scrape user reviews from the merchant site (or Trustpilot). Look for keywords that the merchant hides but customers reveal.

• Customer says: "Cancelled my cable subscription thanks to this!"
• Risk Signal: Device enables copyright infringement.

2. Dispute Text Mining

Analyze chargeback reason codes and comments.

• Signal: "Item Not As Described" disputes where comments mention "buffering", "login failed", "subscription expired". These are software/service issues, not hardware issues.

3. Defensive Pseudocode (Risk Scoring)

// TypeScript: Heuristic to detect "Loaded Hardware" risk
function calculateProductRisk(productDescription: string, userReviews: string[]): number {
    let riskScore = 0;
    
    // Evasive merchant keywords
    const evasionKeywords = ['unlocked', 'jailbroken', 'loaded', 'free tv'];
    
    // Revealing customer keywords
    const customerKeywords = ['channels', 'buffering', 'subscription', 'cable'];

    if (evasionKeywords.some(kw => productDescription.toLowerCase().includes(kw))) {
        riskScore += 30;
    }

    // Check reviews for truth
    const suspiciousReviews = userReviews.filter(review => 
        customerKeywords.some(kw => review.toLowerCase().includes(kw))
    );

    if (suspiciousReviews.length > 0) {
        riskScore += 50; // High confidence signal
    }

    return riskScore;
}