The Quiet Challenge to Stripe’s Risk Layer
Stripe Radar has long been the default fraud-prevention layer for startups and mid-market companies running on Stripe’s payment rails. It works, it’s convenient, and it requires almost no configuration to get started. But convenience has a ceiling, and a growing number of fintechs and neobanks are hitting it – particularly companies that need fraud signals beyond the transaction itself.
That’s where Sardine is carving out real ground.
Sardine’s fraud and compliance API is built on a fundamentally different premise than Radar. Rather than scoring transactions in isolation, Sardine ingests behavioral data – device telemetry, session patterns, typing cadence, scroll behavior – and combines that with financial signals to produce risk scores that travel with the user across their entire journey, not just at checkout. For platforms dealing with account takeovers, synthetic identity fraud, and first-party fraud, this architecture is a material advantage over a system optimized for card-present risk.
Where Radar’s Architecture Falls Short
Stripe Radar is exceptional at what it was built to do: evaluate payment transactions in real time using a network of card data, BIN intelligence, and merchant-specific rules. For an e-commerce company processing standard card payments, it handles the heavy lifting without requiring a dedicated risk team. But financial services companies – crypto exchanges, buy-now-pay-later lenders, digital banks – face a different threat profile. Their fraud isn’t primarily card fraud. It’s identity fraud at onboarding, mule account behavior, and behavioral anomalies that precede a fraudulent transfer by days or weeks.
Radar has no native answer for that gap. It doesn’t score onboarding events, it doesn’t track behavioral signals at the session level, and it doesn’t produce SAR (Suspicious Activity Report) workflows for AML compliance. Companies building on Stripe often end up layering third-party identity verification tools, separate device fingerprinting vendors, and manual compliance queues – a patchwork that creates latency, data silos, and operational overhead. Sardine collapses that stack into one API surface, which is a significant operational argument when you’re trying to scale a compliance function without scaling headcount at the same rate.
The behavioral intelligence layer is where Sardine separates itself most clearly. The platform analyzes how a user interacts with a form – hesitation patterns, copy-paste behavior, mouse movements – to detect scripted automation and synthetic identity signals before a single dollar moves. This isn’t a new concept in enterprise fraud; banks have used behavioral biometrics for years. What Sardine is doing is packaging it in a way that’s accessible to startups and growth-stage fintechs that previously couldn’t afford the enterprise contracts required to access that capability.
The Competitive Math for Fintechs
For a fintech company already running on Stripe, switching away from Radar isn’t cost-free. There’s integration work, rule migration, and the loss of Radar’s native feedback loop with Stripe’s transaction data. But for companies that aren’t exclusively processing card payments – or that have payment infrastructure spread across multiple processors – the lock-in argument for Radar weakens considerably. Sardine’s API is processor-agnostic, which means it works whether a company is running on Stripe, Adyen, or a direct bank connection. That flexibility matters more as fintechs mature and begin routing transactions across multiple rails for cost or redundancy reasons.
Pricing is another dimension where Sardine competes effectively. Radar charges a per-transaction fee on top of Stripe’s standard processing costs, which is manageable at low volumes but compounds at scale. Sardine’s pricing model is structured around API calls and risk decisions rather than transaction volume, which can produce meaningful cost differences for high-volume platforms with relatively low fraud rates – platforms that are effectively subsidizing Radar’s network costs without seeing proportional benefit. The actual cost differential depends heavily on a company’s transaction mix and fraud profile, but the structural difference in how the cost is calculated gives fintechs a reason to run the numbers.
What accelerates Sardine’s traction is regulatory pressure. Financial companies operating under FinCEN guidelines, state money transmitter licenses, or banking-as-a-service partnerships with chartered banks face compliance requirements that go well beyond what Radar addresses. Sardine built its platform to produce compliance-ready outputs from the start – risk scores that map to BSA/AML obligations, audit trails that satisfy examiner requests, and case management tooling for compliance analysts. That positions it not as a fraud tool that compliance can tolerate, but as a compliance infrastructure layer that also handles fraud, which is a more defensible procurement story inside a regulated company.
The Stakes for Stripe’s Risk Business
Stripe has obvious paths to respond – acquiring behavioral biometrics capability, deepening Radar’s AML tooling, or building out onboarding risk scoring through its existing identity products. But the window for those moves to arrive before Sardine establishes deeper roots in the fintech stack is narrowing, and every compliance team that standardizes on Sardine’s case management workflow becomes a harder migration target, regardless of what Stripe ships next.
