Embrace the complexity: How Dutch startup Floryn makes B2B lending fast and efficient

For generations, banking as an industry has always been seen as an operation that is slow and dependent on legacy systems. That nature of business doesn’t change even for business customers. This slow process adversely affects startups and entrepreneurs seeking capital or funding.

According to CB Insights, 38 per cent of startup failure can be owed to failure to raise new capital or running out of cash. Can software and technology solve this cash flow/funding problem for startups as well as small and medium enterprises? Floryn, a Dutch fintech startup based in ‘s-Hertogenbosch, has a novel solution.

Is money on the table?

Floryn was started by Gion van den Bogaert, Marijn van Aerle, and Sven van der Biezen in May 2016. According to Marijn, the biggest hurdle faced by entrepreneurs is knowing “whether the money is on the table or not.”

He says many banks can take several weeks or months to offer clarity on whether they are going to offer the funding sought by entrepreneurs. “This is frustrating and puts your business in waiting mode,” he adds.

Before starting Floryn, Marijn and Sven came from the experience of having sold their previous startup to an American company. They left the firm after three years and spent a lot of time talking to people facing problems that could be solved with software.

The idea of Floryn really came to fruition when they met Gion, who became the third co-founder of Floryn. “He showed us how lending worked at a major Dutch bank and we saw a lot of opportunities there,” explains Marijn.

In simple terms, banks are behind due to use of legacy systems and little focus on the customer. Gion showed Marijn and Sven how they rely on complicated human workflows. All this meant there was a large volume of data available and the ability to automate things.

“It ticked two important boxes for us. It’s a complicated problem and we believe we can make it at least 5x better for customers with software,” Marijn adds.

An AI and software-based fast B2B lending platform

A look into traditional banking systems’ slow approach and belief that software can solve it led to formation of Floryn. At Floryn, Marijn says they make the credit application process fast, easy and transparent.

“We do this with a really easy to use platform that the customer can use to track their application and submit information. Our PSD2 licence allows us to directly connect to the bank account of the business and read out their data,” he illustrates.

Once the data is read, Floryn combines AI models with human underwriters to not only make the right decision but do it fast. “Apply in 2 minutes, arranged within 24 hours” is the first line you will read on Floryn’s website.

The Dutch startup prides on the speed at which credit is offered to businesses. A business applying for a loan on Floryn will know whether they will get it within a day. The fastest conversion, according to Marijn, was 4 hours and 22 minutes from signing up on their website to transfer of money.

Time to check credit worthiness

That speed and quick response is possible because of extreme focus on customer experience. Marijn says they are obsessed when it comes to customer experience and take everything seriously, including the sign up process or having a great mobile app. However, the part that is most important in the whole process is “making the credit decision.”

The process begins with the customer connecting their bank account to the platform. Once connected, Floryn is able to read all the transactions and even consolidate multiple accounts from different banks together. The AI-based classification algorithm then labels each transaction with tags such as revenue, taxes, financing, et cetera.

Marijn says, “We have a couple of major categories and more than hundred subcategories that we label from the transactions. Once we have these labels, we can start aggregating the transaction amount and find relationships between them.”

Once all the data is available, Floryn sends that to its machine learning models and runs different versions and iterations. The ML model turns the data into an easy to use dashboard for its underwriters and sales people. This model also provides guidance to underwriters on easy decisions leading to rejection and approvals.

“The model also provides consistency in the decision making over time, while the humans can provide feedback when circumstances change rapidly, such as when COVID first hit us. The combination of humans and models, and underwriters and data scientists together really makes it work,” Marijn adds.

In addition to this data-driven approach, Floryn also looks for businesses that can grow and claims to have “low loss rate” despite being on the risky side of lending. In terms of approval, Floryn tends to look at businesses with at least €100k in yearly revenue and offers loans ranging from €10k up to €3M.

A unique business model

Once the loan is approved, Floryn directly credits the money to the account of the business. However, the business can decide to withdraw the money at any point in time. Floryn also allows them to repay the loan at any point in time without penalties.

According to Crunchbase, Floryn has raised €78.4M so far and has a simple fee structure. The businesses borrowing money are only required to pay interest on the money they have borrowed. If a business seeks flexibility and pays only for the money they have used then they are required to pay a small withdrawal fee.

However, if a business sees continuous usage then they can pay a quarterly fee instead of the drawdown fee. “Our interest rates are very competitive compared to alternative lenders, and as we’ve grown, we’ve reduced our rates over time, since our costs of borrowing have also gone down,” Marijn says.

With strong growth in the high volume loan segment, Marijn says Floryn is “seeing a lot of growth opportunities within the Netherlands.” He says the market is far from saturated and is planning to diversify its lending product to increase the addressable market. “Our ambitions have always been bigger than The Netherlands,” he says but does not offer any plans for international expansion.

AWS to power the lending ambition

Floryn wants to be the alternative to traditional lending mechanisms and one that relies heavily on software and scale. To reach that scale and fast growth, the Dutch startup relies on AWS for a complete cloud-based solution.

Marijn says, “We still don’t have any physical servers anywhere in the company. I think this is basically the standard for startups nowadays.”

He says AWS offered the most mature cloud solution when they began in 2016 with very broad tools and support. Another factor that helped them choose AWS was endorsement from the Dutch Central Bank approving the use of AWS public cloud service for financial services.

A number of fintech startups rely extensively on Amazon Web Services for their cloud infrastructure. “AWS allows us to scale fast, be flexible when choosing tools and technologies, have nice developer tools and at the same time be compliant with regulatory requirements,” Marijn adds.

“There are a lot of tools included with AWS such as AWS Config, GuardDuty, Security Hub, WAF, Cloudtrail, and things like the Well Architected Framework that help you with having control over your environment on a level required by auditors, while not slowing down your technical staff,” he elaborates.

Embrace the complexity

For Marijn and his co-founders, the barometer of success is not only proportional to the amount of money being lended by the platform. They also want to be seen as a platform that supports entrepreneurs and makes them happy. “We’ve demonstrated that by being very personal, transparent and by removing the paperwork, we can actually make people happy during the lending process,” Marijn says.

In order to further this mission, Floryn is looking for talent in the field of software engineering, data analytics, and machine learning. The startup was the 9th fastest growing company at the Deloitte Fast 50 last year with a growth of 1,153 per cent.

This success, according to Marijn, has come from their embrace of complexity associated with fintech. “The need to deal with money, regulators, auditors, legal, tech, legacy systems, etc broadens your skill set,” he says.

“We built our own payment and accounting system from scratch. I’ve also met people who’ve written their own PSD2 licence and are successful,” he says. “These topics are hard but they’re not magical. For me, this is what makes a Fintech company fun, there are new and interesting puzzles to solve every day,”