Alex Johnson noticed a flurry of startups tackling this problem as he monitored fundraising announcements for fintechs.
"There have been an unusual number of U.S.-based fintech startups that have raised money specifically around solving a KYB issue," said Johnson, author of the Fintech Takes newsletter, including Baselayer, Ballerine, Coris, TrueBiz, Parcha, Accend and Greenlite. It could indicate a "groupthink" mentality in fintech, when "different people become obsessed with the same idea," he said. But it also suggests there is a gap in the market where existing solutions aren't cutting it.
There are numerous reasons why the KYB space is hot right now — and why banks are eager for more automated solutions to ensure companies are legitimate and not shells, to understand who the beneficial owners are and to screen for sanctions and adverse media. There were more than 430,000 business applications in the U.S. in June, according to data from the U.S. Census Bureau — about double the number of applications ten years ago. "It's never been easier to set up a digital storefront," said Danny Hakimian, CEO and founder of TrueBiz, which automates the process of investigating a business's web presence.
Verifying the identity of a business is a much more complex task than verifying the identity of an individual, yet the underlying technology used in both is the same.
KYB is also a highly manual process.
"There has been a huge push to get the KYC process in general more efficient, inclusive of KYB, because it's a massive drain on bank spend," said Spencer Schulten, executive director and U.S. head of financial crimes compliance for Capco. "Non-U.S. customers from higher-risk jurisdictions are seeking access to U.S. banks at a higher rate than I've seen in some time." At the same time, compliance budgets are not growing to keep up with the demand, meaning fewer people have less time to do a greater amount of work.
The complexity of KYB is also becoming easier to solve.
"State business registries weren't online 15 years ago, so you couldn't build a direct API integration to pull data from them before," said Johnson.
'We do a lot of Googling'
First Internet Bank in Fishers, Indiana, and
Anne Sharkey, chief risk officer at the $5.3 billion-asset
Normally, "we do a lot of Googling," she said. "Do they have a website? Does the website work? Does it represent what the business says it does? Are there negative reviews out there about the company or owners?"
She sees potential in AI for flagging negative reviews, or a business model that First Internet does not want to bank (for instance, one relating to marijuana), or a nonfunctioning website, which could indicate the company is fraudulent.
"It still takes human intervention but making the process more efficient allows you to hone in on something that may look irregular, rather than getting buried in the detail and missing something right in front of you," said Sharkey.
While Nbkc uses Middesk to retrieve Secretary of State data, not all states make this information available to the public. That means business owners need to upload such documents manually for an Nbkc employee to review. The $1.2 billion-asset bank is testing a homegrown system it built using Optical Character Recognition, or OCR, to extract the data it needs for its records and to review for certain
Nbkc had explored using a vendor, "but we weren't sold on any in particular," either because of pricing or functionality, said Brian Fellows, director of risk management at Nbkc. The bank will debut the technology this month.
The hope is that it will pare down the work for bank employees and reduce costs.
"Most of the time if [businesses] apply with us and we take a day or two to approve them, they've gone to another bank to open an account," said Fellows.
He has explored using other startups to fill gaps in the bank's current KYB process, but so far hasn't found a combination that beats his current setup and that would justify the price of layering on more solutions.
What sets KYB fintechs apart
In his survey of the landscape, Johnson has wondered what distinguishes the slate of startups he has tracked from each other and from legacy providers.
"Why does the market need 12 of you?" he said.
Hakimian thinks there is room for venture-scale companies that focus on different angles to emerge in this space rather than a winner-takes-all. TrueBiz is not an end-to-end provider.
"Business entity verification is multifaceted and complex," he said.
The 3-year-old TrueBiz scours a business's website, social media, online reviews and more. Clients can use its API to retrieve a report analyzing hundreds of data points to construct a picture of the business's owner and employees, the types of product it offers, its reputation, likelihood of fraud and more, in under 30 seconds, according to Hakimian. TrueBiz's configurable decision modeling layer also indicates whether further due diligence is needed. Its clients tend to be acquiring banks in the U.S. and global fintechs offering merchant services.
Others say they are the rare or sole provider of certain capabilities.
Baselayer tracks the "pulls" on a particular business's information, as if a lender was pulling an individual's credit report. Numerous pulls may be a sign of synthetic identity fraud, where criminals obtain personally identifiable information on different individuals and entities, and reformulate them into multiple fraudulent applications across banks. Baselayer also tracks the businesses and their repayment performance throughout its network to enhance risk profiling.
Co-founder and CEO Jonathan Awad says his is one of a handful of companies in the market that have purchased Secretary of State data, and that his is the only one tracking pulls on the entire profile of a business.
"Our solution is focused on the timeframe from when you get an application to every check you need to get through for Customer Identification Program or underwriting purposes," said Awad.
