Paystack

Shola Akunlade had been building software since college in Lagos, and the problem that became Paystack was already visible in his work. Before Paystack, a business in Nigeria could take three or four weeks just to start accepting payments online. Customers had their own problems too. For many of them, the first time they used a bank card online was on a Paystack checkout page. Shola did not think the problem was too hard to solve. He thought it was simply that nobody had built simple APIs for it.

The first version was rough. Shola could charge a card from his own computer before he had a company, and that convinced him the problem was solvable. He pushed hard to ship, and he has said the first live transaction came about eight months after he started working on it in January. Co-founder Ezra, a former classmate and an engineer who had tried payments before, joined to build a Node plugin, since Shola thought developers would want plugins. Settlement was handled manually at first: a team member reviewed successful transactions and paid merchants by hand. Shola said it was "very crappy" but it let people accept payments. Compliance meant working through PCI DSS, and he split his time roughly into thirds: infrastructure, code, and talking to people.

Paystack's first customers came through personal relationships. Shola built a waitlist of about 300 people by speaking to them directly, and he deliberately kept access limited, which made it feel exclusive and got people talking. The first paying customer was a fruit delivery business. Shola built its website and payment flow himself, and it went live. More customers followed. He also talked to banks and startups that had payment needs, and he was building a reputation for solving them quickly. A friend's email to Y Combinator's partners, and a long, carefully written reply from Shola, led to his application. Paystack was accepted into Y Combinator. The company went on to add co-founder Ezra and a designer, and the team grew from two to five by the end of the batch.

Shola said the early distribution strategy worked because it started with smaller companies whose friends were the bigger companies, so "the sales conversations were really easy" once a large prospect heard about Paystack from a friend or relative. Focus also mattered: Shola turned down easier, custom-built side projects, including an expense management product, so he could stay on the payments problem. Some things did not work. The original checkout form was closely modeled on Stripe Checkout, and it confused users who were not used to entering card details that way. The team reworked it for the customers they were actually serving. Shola also spent too long worrying about large companies switching over, when the smaller ones were the real path to growth.

By the time of the interview, the company said it was processing about $26 million a month, which Shola described as a huge jump from the roughly $2,500 processed in its first month. The team had grown to about 36 people. Shola said the company reached product-market fit only recently, when it was getting the customers it wanted and they were finding the product easy to understand.