Netcompany Raises the Top Line, Holds the Margin, and Sends the Cash Flow the Other Way

Netcompany's Q3 2026 print is a study in a company that is winning the right deals but not yet converting them into cash. Group revenue came in at DKK 2.3bn, up 7.6% in constant currencies excluding Netcompany Banking Services and 8.2% as reported after a 0.6pp currency drag, with cumulative nine-month growth above 28% including the acquired banking unit and 12.6% constant-currency organic ex-NBS. The headline win was NHS England's adult screening programmes — a five-year, GBP 250m contract starting September 2026 covering all seven national programmes, with a new national breast screening service and an enhanced cervical service including at-home HPV self-testing. The U.K. segment grew 42%, with public sector up 50.8%. Denmark grew just 2.6%, the Continent unit 4.8%, Norway was flat and the Netherlands 15%.

The raise that was really a re-basing

Management raised the floor of full-year revenue guidance by two points — to 8.5–10.5% ex-NBS and 18–20.5% for the group — and left the margin band exactly where it was: 17–20% ex-NBS, roughly 16–19% group. That asymmetry is the whole story of the quarter. To reach even the middle of the margin range, Q4 has to carry a step-up, and the CFO was explicit that the lever is revenue mix, not cost.

And the answer to that is not going to be as detailed as you would require, Claus. I think you can appreciate why... it's solely depending on mix.

He added that there are “a number of large projects in our pipeline that have a significant amount of licenses” — Thomas Johansen, Chief Financial Officer (CFO) · 2026-10-29 — but declined to discuss timing. This is not a new pattern. In August, George Webb asked essentially the same question about the implied second-half margin pickup, and got essentially the same answer, that lower license realisation in the first half would reverse. The difference now is that the year has been consumed: group adjusted EBITDA margin before allocated headquarter cost rose to 18.9% from 17.3%, Denmark jumped 11.5pp sequentially to 27.8% (though down 1.6pp year-on-year on marketing spend), the U.K. reached 15%, the Netherlands 24.6% and Banking Services 12%, up 5.6pp on synergies. Norway sat at negative 2.1%. The margin story is being carried by everything except the licensing mix that the guidance depends on.

Sovereign AI: the thesis is old, the wrapper is new

The prepared remarks pivoted hard to Europe's Cloud and AI Development Act and its four union assurance levels, with Netcompany claiming a leading position across all four tiers of sovereign AI for regulated businesses. This is the quarter's freshest framing — the keyword trajectory shows sovereignty and AI Development Act language surging into the company's top themes for the first time. But the underlying claim is not new. In February the CEO was already describing dialogues with governments “affected by the whole movement towards more resilient and a much more strategically independent solutions in the EU space” — André Rogaczewski, CEO and Co-Founder · 2026-02-03, and in August he was arguing that European customers “don't want to use global LLMs where they don't know how much it's going to cost in 1 or 2 years' time” — André Rogaczewski, CEO · 2026-08-13. What has changed is that the pitch now has a regulatory hook (CADA), a named orchestration product — Phoenix AI, apparently a rebrand of the Feniks AI that was a top-ten company keyword in Q2 — and a delivery proof point in the NHS.

I think everything that customers are talking about right now is basically agents, agents and even more agents.

The AI agent theme is genuinely crowded across this week's reporters — consumer agents at Twilio and DoorDash, agentic commerce at Adyen, superintelligence control planes at ServiceNow, agentic security at Tenable — so Netcompany is riding a wave, not creating one. The interesting divergence is one layer down: the global tape has been voting against the infrastructure expression of this theme. The broad AI data center basket is up on a 360-day view (44 advancing tickers versus 12 declining) but down roughly 5% over the last 90 days, with high-bandwidth memory, co-packaged optics and design-win momentum all in the 90-day decliners. Netcompany sells orchestration and modernisation of old systems, not the capex layer — which insulates it from that specific unwind, but also means its agent narrative will be judged on delivery, not on multiple expansion.

The cash bill, and the Denmark promise that keeps slipping

The quarter's ugliest number is free cash flow: negative DKK 170.6m against negative DKK 55.5m a year earlier. The CFO laid the blame on a year-to-date working capital drag of roughly DKK 950m, more than half of it from the disappearance of prepayment from the former owners of SDC — cash that used to be front-loaded into the banking central now simply does not exist in a commercial customer relationship — plus about DKK 300m of RRF-funded receivables collected in Q4 rather than Q3. Days sales outstanding went from 53 to 61, and would have been flat ex-RRF. The drag on working capital is finally being asked about in the same breath as cash conversion, a theme that has now resurfaced in two of the last four quarters after management promised in February that 2026 would look “more in line with what we've seen from a cash conversion perspective like 2025” — Thomas Johansen, CFO · 2026-02-03.

Then there is Denmark. Revenue growth of 2.6% was blamed on ~150 client-facing FTEs moved into the banking segment (−4pp) plus weaker Germany and Sweden (−3pp), with the CEO calling himself “slightly positive on the public sector” — André Rogaczewski, Chief Executive Officer (CEO) and Co-Founder · 2026-10-29 and pointing to bids to be won in Q4 and Q1. Compare that with August's promise, when the same question about Danish public sector conversion drew: “it's been a slow first half year, but we expect it to pick up over the second half year” — André Rogaczewski, CEO · 2026-08-13, subsequently pinned to “second half without being specific” — Thomas Johansen, CFO · 2026-08-13. We are now in the second half, and the pickup is being deferred again. Meanwhile the segment that dominated the company's keyword profile for three straight quarters — Banking — has quietly become a supporting act: integration is delivering, synergies are on plan, and management no longer leads with it.

The investment case therefore rests on one unmoved dial. If Q4 licenses land, the revenue raise is real, the margin band is reachable and the working capital unwinds enough to keep 2026 from being a cash write-off. If they slip, Netcompany will have delivered 8% growth and a genuine marquee contract while still explaining, in February, why cash conversion disappointed. The agents and the sovereignty story are compelling; the invoice is still in the post.