When a construction management platform starts offering invoice financing to its subcontractor network, it is not becoming a financial services company. It is using financial products as a retention and monetisation mechanism within a context where it already has superior information — contract values, payment schedules, completion milestones — than any bank could acquire through a credit application. This is the embedded finance thesis stated concretely: financial products deliver more value, and can be priced more accurately, when they are offered at the point of commercial activity rather than as standalone products sold by institutions with limited context about the customer's actual business.
The infrastructure enabling this shift exists today in a way it did not five years ago. An e-money institution licence, once the province of specialist payments lawyers and six-figure compliance budgets, can now be obtained as a service — through programme managers and Banking-as-a-Service providers who handle the regulatory authorisation, the scheme membership, and the capital requirements. A software platform wanting to offer a stored-value wallet or a card product does not need to become an FCA-authorised firm; it needs an API integration and a commercial agreement. The regulatory complexity has not disappeared, but it has been abstracted. That abstraction is itself a business — and it is where we see the infrastructure opportunity.
There is a legitimate question about where the durable value sits in this stack. A platform that embeds financial products via a BaaS provider is exposed to that provider's pricing power, reliability, and regulatory standing. If the BaaS layer consolidates — which we expect it will — the margin available to the platform above it compresses. The companies we find interesting are those building proprietary financial product capabilities at the infrastructure layer itself: the orchestration APIs that sit between multiple BaaS providers and the platform above, the credit decisioning engines that give the platform genuine underwriting ownership, the reconciliation and treasury logic that makes multi-currency float management tractable for a small operations team.
The historical analogy that feels most apt is the emergence of cloud infrastructure in the early 2010s. Every software company did not become a data centre operator — they used AWS or Azure as infrastructure and competed on the application layer above it. But the companies that built the tooling, monitoring, and orchestration layers between applications and cloud infrastructure — the monitoring and infrastructure-as-code equivalents — accrued significant value precisely because they sat at the seam between general-purpose infrastructure and specific application needs. Embedded finance will produce the same layer. We are investing at that seam.
What we look for specifically is a platform with a clear answer to the question of customer data ownership. The most defensible embedded finance propositions are those where the software platform has transaction-level visibility that cannot be replicated by a bank or a standalone financial product. A fleet management platform that sees fuel spend, maintenance costs, and driver behaviour has a fundamentally different credit perspective on a trucking company than HSBC's relationship manager. That information asymmetry — if converted into a financial product with appropriate regulatory structure — is a durable advantage. We are not interested in embedded finance that repackages existing bank products without a genuine data or distribution advantage. The structural edge has to be real.