Organic growth improves to -1.5% and the margin ticks up, but a 2,000–3,000 headcount ambition lands while the price-versus-salary spread is still flat — and Places keeps promising the same data-center recovery.
AFRY.ST · Earnings Call · 2026-10-22
The pivot finally has a number
For four straight quarters the AFRY story was subtraction: an extensive restructuring agenda, capacity adjustments, structural take-outs rolling through 2025 and into 2026. In Q3 2026 management finally put a number on the other side of the ledger. Linda Palsson committed to “a net growth of 2,000 to 3,000 employees by the end of 2028” — Linda Palsson, President and CEO · 2026-10-22 — roughly 15% more people than today, a figure Dan Heimer of SEB immediately read as an implicit demand call. Organic growth was still -1.5%, but that was the least negative since Q1 2025, and total growth turned positive at +1.9%, carried by the AMC acquisition. Order backlog stood at SEK 22.1 billion, +8% year over year, with every global division up — the same rate as Q2, and therefore a plateau rather than an acceleration.
That hiring target is a genuine change of register. In July, Palsson's answer to the identical recruitment question was directional and unquantified: she said AFRY was “shifting focus even more on the organic growth part” — Linda Pålsson, CEO · 2026-07-15. Three months later the ambition has a headcount range and an end-date. Whether that is confidence or a target set because the 2028 roadmap required one is the question this quarter does not settle.
The awkward part: the spread isn't fixed
What makes the pivot uncomfortable is that AFRY is hiring into a pricing environment it has not yet turned. Jesper Stugemo asked directly whether the company still earns positive yield between price and salary increases. The CFO's answer was as blunt as these things get:
When we closed Q2, we were not actually in the point where we had a clear positive spread on that yield. And I would say that nothing has changed from a Q3 perspective on that end.
Two quarters ago the same weakness was framed as volume, not price: “the vast majority of that is volume driven. We don't have a positive effect from pricing in the quarter” — Bo Sandström, CFO · 2026-07-15. So price pressure is now a multi-quarter refrain that has narrowed rather than disappeared — automotive on the industrial side, places on the real estate side. Palsson was careful this quarter to say the Places signals are “not yet fully materialized” — Linda Palsson, President and CEO · 2026-10-22.
Where the margin actually came from
EBITA margin of 6.8% was up 0.4 percentage points, and utilization of 72.9% marked a fourth consecutive year-over-year improvement, creeping toward the 74% full-year ambition embedded in the 2028 plan. That is the good news and it is real: utilization rate improvement has been the one KPI that has moved in a straight line through the whole restructuring.
The margin, though, was explicitly two things. Industry's improvement of more than a percentage point came from the acquisition of AMC — an entity Bo Sandstrom described as “typically operating on between 10% and 15% EBITDA margin” — Bo Sandstrom, CFO or Financial Officer · 2026-10-22 — plus positive project closings worth SEK 10 million to SEK 20 million, which he conceded carry more of a one-off characteristic. Julia Sundvall's request to split one-off from run-rate got a range, not a reconciliation. That is the same question Dan Heimer asked in July about slack in the system; the answer has simply migrated from "suppressed organic growth" to "AMC and project closings."
Cash was the softer spot. Operating cash flow ran SEK 350 million below last year, attributed to advanced payments — and Bo pushed back on the idea that it reverses: “the advanced payment perspective is more related to a quite high level of advanced payments in Q3 last year” — Bo Sandstrom, CFO or Financial Officer · 2026-10-22. Net debt rose to SEK 4.8 billion, leverage held at 3.0x, with the 2.5x year-end target leaning on normal Q4 seasonality plus restructuring costs rolling out of the trailing EBITDA base.
Places is riding a wave that has already broken
The most interesting cross-read is Transportation & Places, where the 500-employee reduction year over year has not stopped a roughly 50 basis point margin decline against a soft real estate market. Management's prescribed cure has been remarkably consistent. In Q2: “going even more into the hospital side of it, going even more into data centers, going more into defense-related buildings” — Linda Pålsson, CEO · 2026-07-15. In Q3: positive signals linked to data centers, hospitals and defense-related buildings. Six months, the same pipeline, still not materialized.
Meanwhile the market's own vote on that theme has cooled. The global AI data center cluster is a large 360-day advancer but a decliner over both 90 and 30 days, with the overwhelming majority of tagged names negative over the last month. AFRY's Places recovery is levered to precisely the capex the tape is currently discounting. Either the SEK 22.1 billion backlog converts into a re-accelerating theme, or the promised catalyst keeps slipping by a quarter. To be fair to management, this is consensus rather than contrarian — data-center demand surfaced across this week's European industrials, from BE and NEX.PA to HTRO.ST, AKZA.AS and even BULTEN.ST's AI infrastructure space. Palsson is riding a wave, not spotting one.
What stopped being said
The word that has quietly left the vocabulary is restructuring. It dominated 2025 and early 2026 and still shows up in the filing language as significant restructuring costs. Now the emphasis is margin development, utilization and recruitment. Bo's line that the effects will phase out in the beginning of next year is doing double duty: it explains the Q4 deleveraging, and it removes the cover that has explained soft margins for four quarters. After Q1 2027 there is no restructuring comparison left to hide behind, and the company will be judged on a price-cost spread it says is not yet improving.
A credible operating turn with a soft underbelly. Utilization and backlog are genuine; the margin quality is mixed; the spread is unproven; and the hiring ambition is being funded ahead of the revenue that would justify it. The next two quarters will reveal whether 2,000–3,000 is a demand read or a roadmap artifact.