Adyen Pawned Its 55% Margin Target to Buy a Longer Runway
The payments platform cut 2026 EBITDA guidance to 50%, scrapped the 2028 margin framework, and asked investors to trust that the money is going into loyalty, billing and agentic rails rather than into a hole.
ADYEN.AS · Earnings Call · 2026-10-28
The trade, stated plainly
Adyen did not have a bad quarter. Net revenue rose 23% to EUR 738 million, processed volume climbed 30% to EUR 451 billion, and Platforms — the smallest but fastest pillar — grew 46% to EUR 100 million with 46 customers now processing over EUR 1 billion a year, up from 31 a year ago. Any payments company on earth would take those numbers.
Then management spent the entire call explaining why it is going to earn less on them. The 2026 EBITDA margin expectation moved to 50%, and the 55%-plus-by-2028 framework — set less than a year ago and reaffirmed as recently as February — was withdrawn outright. “We are, as we mentioned, discontinuing the EBITDA, the hard EBITDA target.” — Hwa Tsao, Interim CFO · 2026-10-28
What replaced it is a story about duration. Ingo Uytdehaage framed the choice as a deliberate refusal to take a margin number that was sitting right there:
If you ask me, could we get there, it would be relatively easy to get to the 55%, but that would be at the risk of longer-term growth.
That is a defensible argument. It is also the third time in roughly twelve months that Adyen has re-cut the shape of its forward numbers, and the analysts on the call knew it. "We've had some kind of guidance resets over the past several years," Bernstein's Harshita Rawat said, asking management to make investors comfortable with the growth story. That question, not the 22% constant-currency print, is the story of this quarter.
The vocabulary changed before the numbers did
The most revealing evidence isn't in the P&L — it's in what management stopped saying. For six straight quarters, Adyen's growth narrative ran on share of wallet: land an enterprise, then take a bigger slice of its volume every year. In October 2025 the CFO was still saying “the majority of volumes within our customer base is to win” — Ethan Tandowsky, Chief Financial Officer · 2025-10-29. Share of wallet was the number-one theme of the Q3 2025 call.
In this quarter's keyword set, share of wallet is simply gone from the top tier. In its place: durability of growth, cost of growth, guidance framework and one-time cost. That is the vocabulary of a company defending a multiple, not one compounding a share-gain machine. When the single biggest driver of your historical model disappears from your own language while a new phrase — cost of growth — appears for the first time, that is the pivot.
And the 55% target had been fenced carefully before. In August, Pieter van der Does described it as descriptive rather than aspirational: “Rather than a target, it's an outcome.” — Pieter van der Does, Co-Founder and Co-CEO · 2026-08-13 In February, Ethan Tandowsky was more concrete: “Our expectation is that we'll grow up to the 55% and above level by 2028.” — Ethan Tandowsky, CFO · 2026-02-13 An outcome that becomes an expectation that becomes a discontinued metric in nine months is a sequence worth flagging.
Note also the cast. This call was led by Uytdehaage and interim CFO Hwa Tsao — the second CFO voice in three quarters, and without the co-CEO who fronted August's Q&A. Continuity of narrative is harder to sell when the narrators keep changing.
Orb, Talon.One, and the price of conviction
The proximate cause is entirely legible. Adyen closed Orb (usage-based billing) and Talon.One (loyalty) in July and August. Together they added about 2 percentage points to Q3 constant-currency growth and 399 of the 5,596 colleagues on payroll. Management now expects 1,000–1,100 total FTE additions this year versus 550–650 on an underlying basis, and it flagged roughly a point of one-time cost for the full company.
Tsao was candid that the thesis did not survive contact with customers intact: “We bought these businesses. They're great businesses, but they are not stand-alone.” — Hwa Tsao, Interim CFO · 2026-10-28 Uytdehaage's version: “It's a capability game.” — Ingo Uytdehaage, Co-CEO · 2026-10-28 Fair — but capability games are paid for with operating leverage, and the loyalty pitch shifts the buyer from the office of the CFO to the marketing department, a genuinely new revenue pool that Uytdehaage leaned on hard.
The awkward contrast is with August, when the same management team was at pains to say this would not become a habit: “We don't have a plan there to be an acquisition machine from now on.” — Pieter van der Does, Co-Founder and Co-CEO · 2026-08-13 Three months later the company is doubling investment into the acquired assets and telling analysts it would "consider M&A again" where time-to-market matters. Not hypocrisy — but the merchant pull argument is doing a lot of work in place of numbers nobody will give, since Orb and Talon.One will not be broken out as a segment.
Agentic commerce, quarter five of early days
Strategically, the interesting claim is that Adyen's rails are pre-built for Agentic commerce. Roughly 10% of volume already sits with a handful of hyperscalers; authentication, fraud decisioning and settlement exist; the company is working with frontier labs on protocols to keep merchants connected to the shopper behind the agent. That is a real, coherent position — and it is also the fifth consecutive quarter in which agentic commerce has been a headline theme while contributing nothing measurable to revenue. Uytdehaage himself conceded “It's still relatively early days for Agentic.” — Ingo Uytdehaage, Co-CEO · 2026-10-28
The tape is not currently paying for that story anywhere. Over the last 90 days the market's agentic and AI-adjacent software cohort has de-rated hard — agentic operating system names are down about 32% over a year, outcome-based pricing models down 22%. Adyen is asking investors to fund an AI-commerce narrative exactly as the market stops funding them generally. The offset is that Adyen's version of the story is not a software fantasy: the 46 platform customers above EUR 1 billion and 312,000 active business customers (up 48%) are real, contracted volume.
What February has to answer
The mechanics of the reset are less dramatic than the framing. Roughly a point of the margin drag is non-recurring integration cost, and stock-based comp is inside the EBITDA definition — Tsao explicitly flagged equity is rising to win talent in "a very competitive marketplace" and that the guide does not back it out. That is a real, permanent dilution of margin quality disguised as a hiring decision.
Meanwhile Q4 is a shrug: full-year 21–23% constant currency, with Q3 at 22%, implies a Q4 spread wide enough to hold anything from a slowdown to a reacceleration, and Tsao would only say the second half should shape up similarly to the first. In February we get 2027 revenue, EBITDA colour and CapEx. What we will not get is the multi-year margin architecture that anchored the equity story since the 2023 investor day.
The most quietly important line on the call may have been the capital allocation one. With S&P's outlook moved to positive and the A- affirmed, Tsao said “capital returns are part of that toolbox.” — Hwa Tsao, Interim CFO · 2026-10-28 Six months ago that was a philosophical maybe. It is now the only lever left that could make "we chose growth over margin" look like capital discipline rather than a blank cheque — and the only one that doesn't require anyone to believe a 2028 number that no longer exists.