Notifications
Clear all
Topic starter
15/08/2026 6:13 pm
TL;DR: E-commerce platforms are being pulled between conversion and fraud control, with ghost stores, first-party abuse, and false declines all exposing the limits of IP, cookie, and rules-based detection, according to Fingerprint. The practical lesson is that buyer trust signals need to be persistent, privacy-conscious, and layered into existing fraud workflows rather than used as a friction-heavy gate.
NHIMG editorial — based on content published by Fingerprint: Device intelligence and checkout fraud in e-commerce
By the numbers:
- Regulators in Australia flagged more than 140 ghost store sites in a single year.
- The Merchant Risk Council’s 2024 report found that merchants reject 6% of all e-commerce orders, but 2 to 10% of those rejected orders are actually legitimate.
- False declines cost $81 billion in lost revenue per year.
Questions worth separating out
Q: What breaks when fraud controls rely on IP addresses and cookies alone?
A: They break when legitimate users change networks, clear browser state, or travel, because the controls confuse normal behaviour for fraud.
Q: Why do ghost stores matter to marketplace security teams?
A: Ghost stores matter because they turn seller impersonation into a trust problem that can damage customers, merchants, and the platform brand at the same time.
Q: How should security teams reduce false declines without weakening fraud controls?
A: Start by separating hard fraud stops from soft operational failures, then improve the context used in payment decisions.
Practitioner guidance
- Implement persistent visitor correlation Use device, network, and behavioural continuity to recognise repeat actors across cookie resets, IP changes, and private browsing sessions.
- Strengthen ghost-store onboarding review Add behavioural and metadata vetting before seller accounts reach active storefront status, so fraudulent merchants are stopped earlier in the lifecycle rather than after customer complaints accumulate.
- Tune fraud decisions by customer journey stage Treat checkout approval, refund handling, and chargeback review as distinct decision points with different evidence thresholds.
What's in the full article
Fingerprint's full article covers the operational detail this post intentionally leaves for the source:
- More detail on how its 100-plus real-time device, network, and behavioural signals are combined into a persistent identifier.
- Examples of where device intelligence can sit alongside AVS, CVV, proxy detection, and 3D Secure without adding unnecessary checkout friction.
- The article's product-level explanation of how repeat visitors remain recognisable after cookie clearing, IP changes, or private browsing.
- The suggested onboarding and fraud review use cases for ghost stores, false declines, and first-party abuse.
👉 Read Fingerprint's analysis of device intelligence for e-commerce fraud control →
Device intelligence for fraud control: are checkout controls keeping up?
Explore further
15/08/2026 6:19 pm
Device intelligence is becoming a trust-control layer, not just a fraud widget. The article shows that fraud teams need continuity across sessions, payments, and onboarding if they want to separate repeat abuse from legitimate return behaviour. That is a governance question, because the quality of the trust signal determines whether the platform protects revenue or drives away customers. Practitioners should treat persistent device identity as one input to a broader trust model.
A question worth separating out:
Q: Who is accountable when fraud prevention damages legitimate customers?
A: Accountability sits across fraud, IAM, product, and customer operations because blocking decisions affect access, revenue, and trust. Teams should define ownership for thresholds, appeals, and recovery paths, and map those responsibilities to risk governance so no single team optimises only for loss reduction at the expense of customer experience.
👉 Read our full editorial: Device intelligence and checkout fraud: balancing trust and conversion
