Adyen has turned its payment processing business into one of Europe’s biggest entrepreneurial success stories. Now it’s quietly deploying artificial intelligence to keep its remarkable growth streak going.
The Amsterdam-based company’s philosophy of building tools from within rather than through acquisitions has been crucial to its rise, from its founding in 2006 to its 2018 IPO. Its approach to AI reflects that same philosophy. The company hasn’t been talking much publicly about its work in this area, but has instead been cautiously deploying its algorithms internally.
Adyen CEO Pieter van der Does told VentureBeat that the company has studiously avoided hype-driven trends and so initially took a conservative approach to AI. But after developing in-house fraud-detection algorithms, van der Does was quickly convinced of AI’s potential.
“When we did our first trials with it, I was surprised how effective it was,” he said.
He spoke to VentureBeat as part of a podcasting partnership with Samsung Next at the Slush technology conference in Helsinki, Finland. The full conversation can be heard here:
As hype around AI began to build, van der Does’ natural reaction was to be dubious. Rather than panicking and rushing to acquire a startup for its technology, the company began building its own algorithms internally. The results almost immediately exceeded expectations. “The benefits of AI are real,” van der Does said.
Adyen is using AI for fraud detection. That has allowed the company to better leverage the large number of data points its system absorbs to make faster and more accurate decisions about people’s creditworthiness. Its risk management service RevenueProtect includes a ShoperDNA feature, which allows for real-time identification of shoppers.
With AI, Adyen has been able to expand the data points and methodology, which in turn enables much more granular examination of a person or pattern when a purchase is being made. Its Risk Engine automates transaction reviews, which reduces review times by 30%, according to the company.
Typically, such fraud detection is measured by how it limits losses. But van der Does said it’s just as important to measure additional transactions that are approved.
He used the example of an Israeli software developer who works in New York and sends a purchase to Europe with a credit card issued in Israel. There’s a good chance that such a transaction might be flagged or blocked. But with AI able to process more detail, it might realize that the developer is just sending flowers and the transaction is probably fine.
This way of incorporating AI fits with Adyen’s solid, problem-solving approach. The company isn’t trying to spin itself as some machine learning-driven wonder. Instead, it treats AI as a practical advance.
“Risk in payments is a way to raise authorization dollars,” he said. “If you allow the transaction, your merchant has more revenue than [they] would have had with the old system. That is good for everyone.”
Adyen was founded 13 years ago by van der Does, who joined with former colleagues from a payment startup called Bibit Global Payment Services. Bibit was acquired by Royal Bank of Scotland in 2004, and van der Does left the financial giant as soon as his lockup agreement expired.
“[Like] many entrepreneurs, I found it frustrating because the speed of execution suddenly is [at the speed of a] large company,” he said. “But I’m actually really grateful for the things I learned there.”
Part of van der Does’ motivation for starting the company was his perception that innovation in payments had slowed following the Bitbit acquisition, as well as eBay’s purchase of PayPal. Rather than just building another interface over payments services for banks and credit cards, Adyen set out to build the entire payment infrastructure.
Initially, this attracted interest from smaller players, like mobile gaming startups that could suddenly offer seamless in-game purchases, as well as ecommerce upstarts like Groupon, whose initial burst of growth helped Adyen go global.
Over time, as the speed and flexibility of the service proved itself, the company began to attract bigger clients, eventually even landing eBay. As Ayden evolved, it also expanded into point-of-sale services for physical retail customers.
But Adyen maintained a conservative approach to growth, which meant the company waited five years before accepting its first outside investment in 2011, though it eventually raised $266 million in venture capital. The first round came mainly from Index Ventures.
Controlling the end-to-end payment infrastructure has allowed Adyen to launch things like its most recent initiative: a platform that allows clients to issue their own virtual and real payment cards. Using Adyen’s API, clients can offer a range of cards targeted at everything from narrow uses to general ongoing employee expense payments, each built on specific limits and rules established by the creators. Adyen further refined the product by observing customers’ challenges, and it was able to implement changes because it manages the whole system.
“I have been called very slow,” said van der Does. “I’ve been told that I run a pension fund for my children, stuff like that. So yes, we got criticized for our pace of execution. I must say, the shareholder who [said this] was very kind, and he said, ‘Pieter, I eat my words — that was wrong.’”