Hiroki Takeuchi and the long build behind GoCardless

Building account-to-account payment infrastructure before the market has a fashionable name for it requires a narrow entry point, patient work on reliability and a credible route into customers' existing software. GoCardless began with direct debit rather than trying to replace every payment method. It made a fragmented bank process accessible through software, then used that operating base to add open-banking products and a wider bank-payment network.

That is the commercial answer in Hiroki Takeuchi's story. Category timing mattered, but endurance came from doing unglamorous infrastructure work while earning trust one integration and one payment flow at a time. The September 2026 completion of Mollie's acquisition of GoCardless gives the story a new endpoint: bank payments now sit beside card and local payment methods in a broader group. It does not prove every early decision was right, but it shows why a specialised rail can remain strategically valuable after the category catches up.

The useful starting point was recurring collection

Takeuchi co-founded GoCardless in 2011. The company's current official profile lists him as co-founder and chief executive. Its original wedge was practical: many smaller organisations needed to collect repeat payments, but direct debit access and administration were cumbersome. A software layer could handle mandates, collection schedules, reconciliation and notifications without requiring each merchant to become a payments specialist.

This was a disciplined starting market. Recurring collections expose a provider to operational detail very quickly. A missed payment affects cash flow. A duplicate collection damages trust. A mandate recorded incorrectly can create both customer-service work and compliance risk. The product therefore had to be more than a clean checkout screen. It had to make a bank scheme predictable enough for membership businesses, subscription companies and invoice-led firms to place inside their normal operations.

The broader founder lesson is that infrastructure categories are often built from a specific workflow, not from the eventual category label. GoCardless's collection documentation distinguishes direct debit from Pay by Bank, while its account-to-account product history records the later open-banking expansion. That broader proposition rests on years spent learning how businesses authorise, initiate, reconcile and recover payments.

The product hid fragmentation rather than wishing it away

Account-to-account payments are not one universal rail. Direct debit schemes, settlement timings, mandate rules, refund rights and required customer communications vary by market. A useful infrastructure provider absorbs some of that difference and presents a more consistent interface to the customer. GoCardless's documentation shows the commercial shape of that work: merchants can use a hosted dashboard, a pre-built partner integration or an API, while the underlying payment scheme still depends on where the payer's bank account is based.

That distinction explains why integrations became part of distribution. Connections to accounting, billing and customer-management products place payment collection inside an existing task. They also create a demanding channel. A platform partner expects stable APIs, reliable reporting and clear responsibility when a transaction fails. The work is less visible than a consumer brand campaign, but it can produce repeatable access to many merchants through one technical relationship.

Regulatory status is part of the product as well. GoCardless states that its UK entity is authorised by the Financial Conduct Authority under the Payment Services Regulations 2017. Authorisation is not an endorsement of commercial performance, but it defines permissions and obligations that a merchant should check. Founders building financial infrastructure cannot treat compliance as a certificate added after the software is finished.

Open banking expanded the job instead of erasing it

Open banking gave account-to-account payments a more visible category and enabled payment initiation with a customer's active approval. GoCardless launched Instant Bank Pay in 2021 to complement bank debit with one-off payments and immediate confirmation. Its current materials distinguish that flow from direct debit: the former can confirm quickly and needs customer authorisation for each payment, while the latter uses a mandate for later collections.

This is a useful example of adding a rail without pretending the earlier one has become obsolete. A business collecting a variable first payment may value rapid confirmation. The same business may still prefer direct debit for a predictable subscription. Product breadth is valuable when the methods solve different stages of the same customer job, not simply because a company can list more features.

GoCardless also acquired open-banking provider Nordigen and developed data and payment products around bank connectivity. Those moves widened the technical surface and the competitive set. They also increased the need to explain consent, coverage, failure handling and differences between markets. Open banking can make bank payments more flexible, but it does not remove local regulation or the operational burden of moving money safely.

London supplied a starting market, not a complete network

London gave GoCardless access to software founders, financial institutions, regulators and a large base of recurring-revenue businesses. The UK also offered an established direct debit scheme and later became an early open-banking market. That combination made the city useful for product development and early commercial conversations.

London was not enough for a global product. Each new market required scheme access, regulatory analysis, banking relationships, support coverage and product localisation. Even a well-designed API cannot make those differences disappear. This is where the infrastructure thesis becomes expensive: the customer wants one integration, while the provider must maintain many connections and operational rulebooks underneath it.

The company response was to build reach directly and through partners. Its present website says the service connects with hundreds of business-software products and reaches thousands of banks for data access. Those are company statements and must be checked before publication. The more durable point is structural: international scale comes from combining a consistent developer experience with local payment competence.

The Mollie combination changes how the build should be judged

Mollie announced its agreement to acquire GoCardless in December 2025 and said the transaction was subject to approvals. The completion announcement dated 1 September 2026 says Koen Köppen leads the combined group, while GoCardless continues as "GoCardless, a Mollie company" under Takeuchi. The official GoCardless profile still describes Takeuchi as co-founder and CEO. Both facts should be rechecked immediately before publication because integration can change reporting lines.

Strategically, the combination joins a bank-payment specialist with a platform known for cards and local European methods. That supports a clear customer proposition: fewer providers and integrations across recurring, card and local payments. Integration risk sits beside that opportunity. Product roadmaps, service levels, contracts and organisational responsibilities can become less clear during a merger. Customers will judge the deal through continuity and useful combined workflows, not through the breadth of the announcement.

For founders, acquisition is not the only valid destination for infrastructure. Here it is evidence that specialisation created an asset another payments company wanted to combine with complementary rails. It is not evidence that all merchants should consolidate with one provider or that bank payments will replace cards.

What another infrastructure founder should copy carefully

The transferable pattern is to choose a painful, repeatable workflow; own the operational complexity; and distribute through systems customers already use. Expansion should follow adjacent customer jobs. In GoCardless's case, recurring collection led to one-off bank payments, account data and a broader network proposition.

What should not be copied blindly is the vocabulary of a global network. A founder should first be able to show scheme coverage, permissions, settlement behaviour, failure processes and support ownership in each market. The key commercial metric is not the number of rails on a slide. It is whether the product reduces payment work or improves collection outcomes without creating new reconciliation and compliance costs.

The long build also depends on governance. Payments infrastructure can scale the consequences of an error as quickly as it scales convenience. Engineering resilience, incident response, safeguarding arrangements and transparent customer communication are core product choices. They are not back-office details.

Limitations and questions still to report

This profile uses public company materials, regulatory references and product documentation. It does not include an interview with Takeuchi, Mollie, former employees, platform partners or merchants. Company-reported customer counts, payment volumes and coverage figures have not been independently audited for this draft and are deliberately not used to claim superior performance.

The article also cannot establish which early customer was the first credible commercial win, how long procurement took, or how unit economics changed across direct debit and open-banking products. Those questions require customer evidence, filings and management response. Before publication, Shoreditch Talk should also confirm the post-acquisition leadership structure, review recent GoCardless accounts and ask customers whether combining payment methods has reduced operational complexity in practice.