Socure Acquires Document Verification Startup Berbix

By Johnny Ayers, Founder and CEO of Socure

Socure Acquires Document Verification Startup Berbix to Smoke the Competition in Speed, Accuracy, and Fraud Capture

A New Era of Document Verification: Socure and Berbix

Stopping Clean True Name Identity Theft

Socure Takes Public Stand to Root out 100,000 Synthetic Identities by Year’s End

Today I am thrilled to announce that <a href="https://www.socure.com/berbix-acquisition-document-verification">Socure has acquired Berbix</a>, a San Francisco-based ID document verification startup that developed a forensics engine which detects spoofed IDs – including AI-generated fakes – that are visually indistinguishable to the human eye. Berbix additionally brings full global coverage with technology that learns new ID documents rapidly in any market worldwide.
The approximately $70 million cash and stock transaction marks the first acquisition for Socure.
With the integration of our technologies already complete, Socure has accelerated its innovation roadmap by at least 18 months. With that, I am equally thrilled to announce the launch of <a href="https://www.socure.com/products/document-verification">Predictive Document Verification (DocV) 3.0</a>, which far outpaces every market competitor in accuracy, speed, user experience, and fraud capture.
For our customers, this radical accuracy and speed means an immediate lift in consumer acquisition and revenue, and a material reduction in fraud losses and related operational spend. 
<h3>Why Berbix? Quantifiable Performance that Smokes the Competition</h3>
There’s an unfortunate but common perception in the market that all document verification solutions perform the same – they are most certainly not. 
I understand why users have lost faith. After testing many ID document verification offerings, we found endless examples of multiple vendors in this space providing inaccurate stats that just couldn’t be validated. In fact, what they actually produced was substantially worse performance in true accepts, true rejects and false accepts, along with much slower responses than what was advertised.
Berbix, however, provided the exact opposite in our testing – they delivered substantially better results than what were promised and they dominated the rest of the market in terms of performance &#8211; and this is why we moved from just partnering to fully acquiring the company. Berbix fits perfectly into the Socure ethos: Underpromise and over deliver and win based on performance, not just because of good marketing. Berbix’s performance in the market also spoke for itself: They had never lost a live head-to-head test against anyone. 
Equally interesting to us, Berbix’s innovation and expertise was distinct and yet complementary to our own. The company was founded in 2018 by Eric Levine and Steve Kirkham, former members of the Airbnb Trust and Safety Team, who focused on a forensics and data extraction engine that reads every human and machine-readable component of an ID (OCRs, MRZs, barcodes) with sophisticated image processing pipelines. The speed and precision in sussing out compromised and fake IDs is second to none.
Meanwhile, the Socure team had focused on a combination of improved user experience and the integration of our unique predictive ID fraud signals and insights to deliver extreme accuracy. 
On the front end, our AI-driven image capture technology identifies images which are blurry or positioned incorrectly, and automatically coaches users through the ID and selfie intake process to reduce drop-off due to friction. 
On the backend, we threaded in predictive machine learning fraud models that are trained on 400+ third-party data sources and leverage our comprehensive graph of 1 billion known good and bad identities with risk signals, including device, phone ownership, behavioral data, address, PII verification and geolocation, among others. 
The combined Socure and Berbix technologies, now culminating in Socure’s DocV 3.0, allow for dozens of AI models to make hundreds of split-second decisions on a consumer and their credentials, resulting in an acceptance or rejection decision in under 4 seconds. This represents a significant departure from legacy providers whose document verification models rely on simple template checks and rules to determine if a document is legitimate or not, with little to no view on the rest of the identity associated with the document. 
Without pairing that documentary check with advanced machine learning models from all other dimensions of the identity, customers will see far less accurate decisions, resulting in higher fraud, and lower customer acceptance. 
This is why we are adamant about telling our DocV 3.0 story through quantifiable metrics and challenge anyone in the market to a head to head test.
<h3>Here’s what DocV 3.0 delivers versus market standards:</h3>
<ul>
<li style="font-weight: 400;" aria-level="1">Fully automated decisioning 14% higher than the market &#8211; DocV 3.0 delivers 94% auto decisioning versus an industry standard of 80%</li>
<li style="font-weight: 400;" aria-level="1">First-attempt auto approvals of good consumers 26% higher than the market  &#8211; DocV 3.0 delivers a 90% approval rate of good consumers versus an industry standard of 64%</li>
<li style="font-weight: 400;" aria-level="1">Fraudulent document capture 27% higher than the market &#8211; DocV 3.0 accurately captures more than 83% of all fraudulent documents versus an industry standard of 56% </li>
<li style="font-weight: 400;" aria-level="1">Speed to decision up to 90% higher than the market &#8211; DocV 3.0 delivers accurate, fully automated conclusive results in less than 4 seconds versus a market range of 6-75 seconds </li>
</ul>
This kind of performance could equate to tens of millions of dollars in revenue and savings resulting from a huge lift in customer acquisition, reduction in fraud losses, and a decrease in the operational spend connected to manual reviews. 
What does that look like in action? A global two-sided ecommerce marketplace, similar to one we have worked with, can see an extra $86.1 million in revenue in a single year with DocV 3.0. 
Here’s the math – 
<ul>
<li style="font-weight: 400;" aria-level="1">If they’re running 500,000 checks a month on consumers with a per-year revenue value of $50, improving auto-acceptance of customers by 26% can mean a lift in revenue of $78 million over just 12 months</li>
<li style="font-weight: 400;" aria-level="1">At the same time, increasing fraud capture by 27% in that same environment where there’s 50 basis points of fraud and an average loss of $1,000, you’d see a return of $8.1 million over the same period</li>
<li style="font-weight: 400;" aria-level="1">That $86.1 million in earned or saved revenue doesn’t include the massive amount of savings accrued by reducing the number of investigators needed to manually review potential incidents or the improvement in your customer’s experience with Socure’s best-in-class first-attempt approval rates and speed to a ‘yes’. </li>
</ul>
<h3>DocV 3.0 Accelerates Innovation in New Markets and Geographic Expansion</h3>
Beyond earned and saved revenue, radical accuracy and speed in DocV 3.0 unlocks innovation in the way our customers can leverage document verification across market verticals and use cases. That’s more important than ever as consumers have brought the most essential – and risky – parts of their lives online, from virtual healthcare to home purchases.
Berbix brings with it a diverse and exciting customer base with interesting use cases in dating sites, alcohol delivery, micromobility, virtual healthcare, and car sharing. In many of these cases, document verification is the only or the first touch in the onboarding workflow. It enables regulatory-required age verification in alcohol delivery and scooter rentals, or ID documentation to confirm drivers in a delivery service or virtual patients on each physician visit. In these cases, a fraud-centric document verification solution could be what prevents an attack by a delivery driver with a felon record who used a synthetic identity to onboard with a car sharing service. And yet, it can also be the answer to seamless, light-touch onboarding.
Meanwhile, at Socure, we’ve seen truly groundbreaking use cases emerge in our own customer base. One of my favorites is a credit union that leverages DocV inside its physical branches, where customers onboard and validate their IDs digitally at a kiosk to bypass the possibility of human error and fake checks at the teller. 
We’re also continuing to see the need for document verification further into the customer lifecycle. Some banks, for example, are using it as a step-up for risky transactions – let’s say before allowing a high value, cross-border wire or sudden account change and then higher-risk transaction. 
As we expand into serving the public sector, fast and accurate document verification is an essential part of the NIST IAL2 flow. Mobile driver&#8217;s licenses and document verification becomes the gateway to delivering essential government services online to every American.
There are a few central elements that unlock this innovation. Clearly speed, customer experience, accuracy, and fraud capture are the support pillars. But flexibility in combining solutions is just as crucial. And that comes from making document verification part of an integrated identity platform along with device, behavioral biometrics, email/phone risk, synthetic &amp; ID fraud, KYC, AML, and bank account validation. That is exactly what we uniquely deliver in our Socure ID+ platform, which allows for the simple orchestration necessary in many of these interesting workflows with best-in-class solutions across every single individual dimension of an ID and risk. 
Finally, all of these use cases are prevalent across the globe – and with the integration of Berbix, we accelerate our pace of expansion into every single new geography. Berbix, which already provides document coverage in 180 nations for ICAO documents such as passports and many local IDs, also brings with it the ability to learn and validate brand new government IDs in any market in the world in under 5 days. This couldn’t be better timing – 60% of our nearly 2,200 customers are leveraging document verification and many are expanding across the globe as they look to onboard food delivery drivers in Mumbai, rent a car in Barcelona, or bank new users in the U.K. We’re thrilled to grow in lockstep with these customers who continue to demand this of Socure and our roadmap. 
<h3>Berbix: An Incredibly Talented Team Perfectly Complements Socure</h3>
Before this acquisition, we had heard the nightmare stories of challenging acquisitions and integrations. But this has been far from the case as the Berbix teams enter Socure’s ranks.
Berbix co-founder Eric Levine, who is now an SVP and Head of Document Verification, is leading a unified team made up of the brightest and most innovative ID verification engineers and designers from Berbix, as well as computer vision, data science, and engineering document verification experts from Socure’s market-defining product team. This group will report directly to our Chief Product &amp; Analytics Officer, Pablo Abreu. 
They share a culture of collaboration, speed, and innovation with a sharp focus on solving customer problems with data science and AI. It has been a complete joy to see these teams gel so easily and quickly. It’s clear that together we are so much stronger, and that the road ahead is open to innovation beyond what we ever imagined in document verification.
Please join us in welcoming Berbix to the Socure team! I look forward to how much more we can build and solve together.

Socure Acquires Document Verification Startup Berbix to Smoke the Competition in Speed, Accuracy, and Fraud Capture

Johnny Ayers

Related Posts