Sodexo's Reset Meets Its Starting-Point Problem
The beat that management immediately un-sells
Sodexo closed fiscal 2026 with revenue near EUR 24 billion, organic growth of +2%, a stronger-than-expected fourth quarter at +2.6%, an underlying operating margin of 3.4% at the upper end of guidance, free cash flow of EUR 379 million, leverage of 2.4x, and a proposed dividend of EUR 1.74. On paper, that is a clean beat-and-confirm. In practice, management spent the call telling investors not to annualize it. Thierry Delaporte called the year one of “diagnosis and reset” — Thierry Delaporte, Chief Executive Officer (CEO) · 2026-10-23, and CFO Sébastien de Tramasure was even more direct about the fourth quarter: “this is not the starting run rate for fiscal year 2027” — Sébastien de Tramasure, Chief Financial Officer (CFO) · 2026-10-23. The starting point is the problem. Q1 organic growth is guided below the full-year 2%-3% range because education turns negative, Sodexo Live! is unlikely to repeat double-digit growth, Meta is still ramping, and a contract reclassification carries a 100 basis point headwind that only drops out in January. For the full year, that reclassification is a 30-35 basis point drag. The year is therefore back-end loaded by design, not by accident — and the market has to underwrite two or three quarters of visible softness before the promised acceleration arrives.Retention is the real KPI, and the market already knew
The headline commercial numbers look better than the guide. Client retention was 94.4% and developments reached 7.7% of revenue. Delaporte framed 94.4% as the third-best retention performance in 15 years, but immediately qualified it: “retention remains below where we want to be” — Thierry Delaporte, Chief Executive Officer (CEO) · 2026-10-23, and versus competitors it is “significantly below” — Thierry Delaporte, Chief Executive Officer (CEO) · 2026-10-23. This is not a new theme. In July, the CFO told analysts retention should land “broadly in line with last year, I would say, around the 94% level” — Sebastien De Tramasure, Chief Financial Officer · 2026-07-02. So 94.4% is a modest beat against a low bar, not a step change. The prior CEO had targeted 95%-96%; the new CEO is now calling the gap structural and deep, which is a more honest framing but also a warning that it will not close quickly.The guidance arithmetic is where the tension sits. A 7.7% development rate and 94.4% retention imply roughly 2% of annualized net new business. Yet FY27 guidance assumes only 0.5%-1% net new, largely driven by Meta. Analysts pressed exactly this point. The CFO's answer was a useful block of honesty:
That is either prudent phasing or a signal that the forward-looking KPI leads reported revenue by more than a year. The language shift is telling. In July, commercial momentum was a top-three company keyword; this quarter it has fallen out of the top 30 entirely. The new top keywords are leadership team, restructuring cost, and asset review. Management has stopped selling momentum and started selling execution. Delaporte made that explicit: “we know that we will never win on strategy. We know that we will win on execution.” — Thierry Delaporte, Chief Executive Officer (CEO) · 2026-10-23We factor a minimum price increase of 2%. In terms of net new, we are expecting a contribution between 0.5% and 1%, and this will be largely driven by Meta... Organic growth is expected... below the range.
The margin bridge is a bridge to 2028
The FY26 margin decline of 130 basis points has three clean buckets: about 50 basis points from the contract and asset review, 30 basis points from higher commercial, supply-chain, and technology operating expenses, and 50 basis points from operational underperformance. FY27 guidance is for a broadly flat 3.4% margin. The positive is that management does not assume a repeat of the review charges. The negative is that the investment step-up and new-contract start-up costs absorb that relief. The asset review is described as largely complete, but the CFO was careful: “our fiscal 2027 outlook does not assume a repeat of the charges, although we will continue to assess contracts and risk with the same level of rigor” — Sébastien de Tramasure, Chief Financial Officer (CFO) · 2026-10-23. In April, management had said of the review that “our objective is precisely to have done the job” — Thierry Delaporte, Chief Executive Officer (CEO) · 2026-04-10. The risk is not a large new charge; it is that the review remains a permanent discipline that can occasionally surprise.Restructuring is the other bridge. EUR 216 million was recognized in FY26 and about EUR 190 million is expected in FY27, keeping the cumulative programme near the EUR 400 million envelope. Cash out was only about EUR 100 million in FY26, with most of the remaining cash impact in FY27. That is why leverage is guided to 2.6x-2.8x, worse than the better-than-expected 2.4x just reported, before deleveraging from FY28. The leverage beat is therefore a timing benefit, not fresh firepower. Management was clear that bolt on M&A is the only realistic use of capital, and that nothing material should be expected. In April, Delaporte had said FY26 guidance could be considered a floor. FY27's flat margin is that floor in practice; the question is whether it becomes a ceiling if Q1 disappoints.
Data centers are a real option, not a new one
The most interesting growth vector is data centers. Delaporte said Sodexo “operate[s] north of 100 data centers” — Thierry Delaporte, Chief Executive Officer (CEO) · 2026-10-23, has built a dedicated offering called Sodexo H, and is leveraging its remote-site management expertise from mining and oil to serve the construction phase, where the value is highest. That is a genuinely differentiated angle. But it is not new: data centers was already a top-two company keyword in July, when the CFO called it a fast-growing opportunity. And the global tape has cooled on the theme — the AI data centers basket is down about 5.6% over the last 90 days after a 52% gain over 360 days. Sodexo is riding a wave the market has already de-rated, though its exposure is services and camp logistics rather than capex. The near-term proof point is Meta: 130 sites across 30 countries, with a normalized margin expected in the second half of FY27. The new protest in France keyword is noise by comparison: France is 12% of the business, the affected school segment is modest, and management sees no material impact.The story is not that Sodexo is broken. It is that the reset is credible but slow, and the market is being asked to pay for execution before it sees it. FY26 was a year of clearing the decks; FY27 is the second year of rebuilding, with the benefits explicitly weighted to FY28. If Q1 marks the trough and retention inflects, the FY28 operating leverage could be sharp. If Q1 comes in below the range and net new stays thin, the floor management pointed to in April becomes the ceiling. For now, the burden of proof is on weekly execution, not strategic ambition.