Payment volume and payment-company revenue are very different quantities. Adyen’s first-half figures make that distinction visible: €803.8 billion of processed volume sits beside €1.303 billion of net revenue. The first is money moving through the platform; the second is the company’s reported revenue measure.
Adyen’s half-year release reported processed-volume growth of 24% and net-revenue growth of 19%, or 21% at constant currencies. EBITDA was €641.5 million, giving a 49% margin, or 50% excluding the disclosed one-time costs. These are first-half results, not full-year outcomes.
More volume does not translate mechanically into more profit
The relationship between processed value and revenue depends on the business being served and the services provided. Two customers can bring similar payment volume and have different commercial arrangements, transaction patterns and costs. That makes a simple volume-growth forecast insufficient for estimating earnings.
Dividing revenue by processed volume would produce an arithmetic ratio, but describing that ratio as a universal customer price would be misleading. A changing mix can move it even when no standard tariff has changed. The better question is whether the platform’s revenue and cost structure improve as customer activity grows.
This distinction also helps separate a payments business from a bank. Both handle money and invest in technology, but their reporting measures and sources of risk differ. ING’s fee-income results provide a useful financial-services comparison without suggesting that the two models are interchangeable.
The acquisitions belong to the next reporting story
Adyen said its Talon.One and Orb acquisitions completed on 1 July. That falls after the first-half reporting period. Their later contribution should therefore not be offered as an explanation for growth in the six months already reported.
An acquisition introduces several questions that an organic growth figure does not answer: how the products connect, which customers adopt the combined offering and what integration costs follow. The purchase decision is only the beginning of that operating evidence.
Future comparisons will need clear labels for acquired and existing activity. Otherwise, readers may attribute a change in consolidated revenue to stronger underlying demand when part of it comes from a different reporting perimeter.
Capacity spending comes before the return can be measured
The company put its 2026 capital-expenditure expectation at 7% of net revenue, reflecting data-centre investment brought forward from 2027. That is guidance about spending and timing, not evidence of an immediate increase in earnings.
Infrastructure can support resilience, capacity and future growth. Its return depends on utilisation and the economics of the business it enables. Spending sooner may be commercially sensible while still reducing near-term cash available for other uses.
DBR’s grid-congestion coverage provides context for the wider Dutch infrastructure debate, but it should not be read as proof that a particular Adyen facility is delayed. Project-specific constraints require project-specific evidence.
The next useful test is whether growth, operating costs and cash conversion remain consistent as the platform expands. The first-half figures establish the starting point; the acquisitions and revised spending timetable mean the second half will require its own careful comparison.