Hi {{first_name|there}},
Most KYB contracts get negotiated on price per verification. Almost none get negotiated on what happens when a verification comes back empty, and that is the part that actually costs money.
In this issue we spotlight how Quantum Lending Solutions used a platform migration to re-test its KYB stack on 2,000 real records, and what moved when the match rate improved.
The cost that isn't on the invoice
The price of a KYB verification is easy to see. It's quoted per call or per verified result, and it sits on a line of the agreement. The cost of a miss is diffuse. It's an analyst opening a file, searching a state registry by hand, deciding whether an address is a real place. It's an application that sits in manual review while these human workflows are happening. Nobody sends an invoice for it, so it tends to survive vendor reviews untouched.
We've been measuring that gap for a while now
When Ramp sent us 4,000 real applications from their production queue, the more revealing finding wasn't the match rate. It was that 87% of the manual reviews in their pipeline were triggered by address verification failures, and that in a sample of roughly 50 flagged addresses, about half turned out to be legitimate physical business locations. The queue wasn't full of fraud. Most of those addresses were real places the verification layer just couldn't match. (That finding is Ramp's, from their own case study, not something we're taking credit for.)
That's the shape of the problem across most onboarding workflows. Match rate improvements don't stay inside the verification step. They propagate. Every point of match rate is a set of cases that never reach an analyst, never wait in a queue, and never cost anyone an afternoon.
Which is why the arithmetic on a KYB switch rarely works the way the contract implies. A ten-point improvement in match rate rarely translates into a ten-point change in API spend. What changes is the workflow queue, and the queue is staffed by people.
The better question for a verification vendor isn't what a verification costs. It's what a miss costs, and how many of those you're already paying for without knowing it.
How Quantum Lending re-tested its KYB stack
Most companies don't reopen a working vendor agreement. There's no trigger for it, and the cost of being wrong is immediate while the benefit is hypothetical.
A platform migration removes that problem. It's one of the few moments when every agreement is on the table at once, and the comparison can run on production data instead of a demo.
Quantum Lending Solutions is an API-first lending infrastructure. Banks, credit unions, fintech originators, non-bank lenders, and CDFIs use it to run SMB lending programs without building the underwriting stack themselves. Every application their platform underwrites requires business verification, so match rates translate directly into how many applications clear automatically. When new investment led to a replatforming, their analytics and risk team used the transition to reassess every vendor agreement, not because the old system was broken, but because replatforming creates a rare window to test alternatives with real data.
They'd been running KYB through a legacy provider for years. Match rates sat around 80% overall, with some verification components as low as 50%. Every missed match meant a manual lookup, and at scale those lookups compound.
They ran a head-to-head comparison on 2,000 real records against three criteria: data quality, cost, and ease of working. Enigma returned a 90% match rate against roughly 80% from the incumbent, at 35% lower direct cost. The downstream effect was 10% fewer applications landing in a manual review queue each month.
3 things decided the switch for their analytics and risk team
Better data cut operational cost, not just API cost: The manual review savings were the multiplier nobody had priced in.
Scalable contract terms: They needed terms sized for where the company is now, with room to scale, not an enterprise-tier minimum built for a different kind of buyer
The integration was easy: Documentation, test records, a test environment that mapped to their existing compliance workflows
If you're re-evaluating a verification stack this year and want to work through what your own miss rate is costing you, get in touch.
Lindsay
Enigma