Today, and especially today, many of us do our banking, taxes, dating, job applications–even doctors appointments–completely online. However, the added convenience of our digitized lives comes with a price. As the amount of highly-sensitive data shared on the Internet grows, it becomes increasingly susceptible and valuable to fraudsters.
Proper identity verification is the first line of defense. Most organizations rely primarily on cumbersome and ineffective manual review processes. This not only imposes high friction for customers, but also exposes businesses and individuals to unnecessary risk from fraud and privacy leaks. By taking humans out of the loop, automated IDV safeguards sensitive information.
What is automated identity verification?
The main goal of identity verification is to make sure that an individual is who they claim to be in any given context:
- Is the driver picking up the same individual who passed a thorough vetting process?
- Is the person trying to purchase goods online the actual owner of the credit card information that is being used?
- Is a fraudster trying to guess your mother’s maiden name to gain access to a bank account?
Automated identity verification solutions streamline the process of validating individuals in trusted transactions by leveraging several technologies: machine learning, computer vision, and automated algorithmic checks. This can take the form of clearing government issued IDs, checking a selfie for liveness and ID match, or cross referencing information across identity databases.
There are major benefits to using automated identity verification solutions instead of manual ones: automated solutions enhance user experience and conversion, enable more comprehensive fraud deterrence, support scale for organizations, and meaningfully increase the security and privacy of users’ sensitive data.
Speed and performance
Depending on the types of verifications performed, manual checks completion times range from several minutes to several days, which can have a brutal impact on conversion.
Accounting for 23% of abandoned sales, drop-off from online processes that take ‘too long or are too complicated’ is the third largest reason for lost e-commerce revenue..
Automated identity verification solutions perform comprehensive checks in a matter of seconds. Persona provides industry leading processing performance with international government issued IDs clocking in at 7 seconds and selfie verifications at 3 seconds.
Asking users to verify themselves during a 20 minute phone call for a stolen credit card is one thing, but to ask users to perform this check for every use case is an impractical, if not impossible, expectation.
Automated Identity platforms can be configured to introduce the appropriate level of friction given any context.
On-boarding a low risk individual, a friend, or known associate? A simple form entry might be enough for verification. Is a high net worth individual signing up for services? Send a more thorough set of inspections. Are fraud signals indicating that an individual from Russia is submitting a California driver license? Add a video selfie verification just for them.
On top of being cumbersome for customers, some manual verifications fail to mitigate risk. Try replacing a lost credit card or resetting a password and organizations will ask users the same knowledge based questions: verify SSN, enter your mother’s maiden name, enter your favorite ice cream flavor. These questions are virtually ineffective in today’s state of identity where personal information is unfortunately easily accessible due to high profile data breaches.
Persona provides the building blocks so that organizations can tailor holistic verification flows where checks are cross-referenced for full assurance:
- International government ID verifications that take 7 seconds
- Video selfies that check liveness and perform ID matching
- Database lookups for validation and KYC
- Reverse phone 2FA and link against aggregate data sources
Passive Fraud Signals
Automated solutions can aggregate fraud signals to make better data driven decisions. Some examples include:
- IP address analysis
- Browser and Device signals
- Proxy or Tor connection detection
- Phone number signals
These are some checks that are tedious to collect from a manual process and impossible to correlate and synthesize in real time for humans.
Bad actors are unlikely to share legitimate PII with any institution that could track or trace them. How can companies prevent a known fraudster from using their services again?
Persona’s Lists feature standardizes inspections that an analyst might have otherwise missed. Companies can define blacklists based on information from IDs or fraud signals such as IP, addresses, or devices, making it harder for fraudsters to evade detection. This kind of detection is unfeasible and almost impossible for manual operations teams to scale.
Is a specific IP address attempting to reset passwords for multiple accounts? Add them to a blacklisted IP List. Do you have high-value customers that should be manually reviewed? For example, is George Clooney a customer? You may want to flag them!
Traditional verification providers require expensive per transaction rates due to fixed headcount. Not only is this a glaring privacy concern, but this presents a high financial burden that customers must shoulder. Customers end up paying high fees for services that sometimes yield no verification result for their business and users.
In stark contrast, automated solutions actually get less expensive as computing volume increases. Additionally, Persona commits to only charging for successful verifications to make sure our incentives are aligned with our customers. This means that if an individual drops off due to user error (blurry images, low-quality camera), Persona customers will not be charged.
Automated solutions simply outperform humans on a per verification basis, but there are other considerations businesses must address when managing manual review teams. Some companies experience spikes in volume either due to seasonality or product virality. On-boarding manual review teams take time to train and headcount costs can’t be scaled back for when verification requests are low.
Online businesses are operational 24/7. Their verification services should be too. Check out our uptime on our Service Status page.
InternationalCoverage and Removing Biases
Verifying all US ID formats is hard enough. Ramping up a team to verify international IDs takes significant resources and training. Organizations need to keep in mind that this is not a one time cost as IDs are constantly updated. Additionally, research has shown that humans are at the mercy of unconscious biases and as a result experience great challenges in recognizing faces of ethnicities that are different from their own.
With so many forms of identification used around the world, machines can process and reference those variations without relying on human judgment. Insights and data collected from these solutions can be used to inform our understanding of global internet fraud so we can continue to hone and improve the solutions we have over time.
Some companies claim to have purported “document experts” to review sensitive information, when in fact these experts are outsourced teams in India or Southeast Asia. Sharing sensitive information offshore in this way raises some serious privacy concerns and questions, as there is little visibility into and accountability for these fleets of offshore contractors.
Reduce Insider Fraud
Insider fraud accounts for 50% of all bank related fraud. Moreover, over 75% of insider fraud originates from functions that include customer support and business operations. Automated solutions limit sharing of PII data on a need to know basis and can compartmentalize and token information for further protection. Persona can track which users had access to information and maintains an Audit Trail in the event an insider fraud investigation is necessary.
Saving the Hardest for the Humans
While automated identity verification solutions will only continue to improve over time, we won’t be automating 100 percent of the decision-making anytime soon. Humans bring a necessary context to the process of identity verification that machines simply do not have — and our ability to judge situations full of nuance and complexity make us indispensable.
At Persona, we know that there is no silver bullet and that automated identity verification can’t solve fraud and privacy problems entirely on its own. Human interaction has an important role in the verification process, helping train and solve edge cases that automated systems and machine learning models may struggle to solve.
The solution? Let your automated service handle the majority of your organization’s day-to-day identity verification tasks, freeing up your security team to deal with the deeply nuanced and complex cases of fraud that require a human and someone who understands your business well.
Reach out to us to learn more about how an automated identity solution can support your business.