Fortinet delivered a quarter that looked strong by almost any operating measure, then got punished as if it had stumbled. For the quarter ended Sept. 30, the cybersecurity vendor reported $2.08 billion in revenue, up 20% year over year, product revenue up 37% to $765 million, and billings up 29% to $2.34 billion. It also raised full-year revenue guidance. Yet on Oct. 29, its shares fell 26% to a five-month low.
That gap is the real story. Security budgets can still be active, products can still be shipping, and cash flow can still be rising even as investors decide the next phase of growth is not predictable enough, recurring enough, or AI-linked enough to deserve the same valuation. For chief information security officers, infrastructure leaders, and procurement teams, the useful question is not whether Fortinet had a “good” quarter. It did. The better question is why a good quarter was not enough, and what that says about how to evaluate security vendors now.
What Fortinet actually reported
In Fortinet’s third-quarter release, the company reported GAAP operating margin of 34% and non-GAAP operating margin of 38%, alongside GAAP earnings per share of $0.86 and non-GAAP EPS of $0.91. Operating cash flow was $946 million and free cash flow was $865 million. Fortinet also increased its share-repurchase authorization by $1.2 billion, leaving $2 billion available.
The full-year outlook moved up as well. Fortinet guided to $8.125 billion to $8.215 billion in 2026 revenue, which the company said implies about 20% year-over-year growth. It also forecast $5.150 billion to $5.170 billion in service revenue and $9.568 billion to $9.718 billion in billings.
That combination matters because it shows more than one kind of strength. Revenue grew. Billings grew faster than revenue. Margins remained high. Cash generation stayed substantial. None of those are signs of a business in visible retreat.
Fortinet also used the release to underline its AI and platform strategy, highlighting FortiGuard Inline AI Security Service for AI-activity visibility and guardrails, FortiLink SASE for distributed sites, and partnerships or selections involving NVIDIA’s OpenShell reference architecture, Oracle, and Microsoft’s sovereign-AI ecosystem. Those announcements are relevant as signals of where the company wants to compete, even if they do not by themselves prove deployment success or measurable customer outcomes.
Why the market still turned sharply negative
The selloff was not a verdict that Fortinet’s products stopped working. It was a repricing of what investors think future growth will look like.
As Reuters reported, investors focused on slower-than-expected product growth and a lower service-revenue forecast. That sounds counterintuitive at first, because 37% product growth is high in absolute terms. But public companies are not graded against zero; they are graded against expectations, and those expectations for established software and security vendors remain unusually demanding.
The service-revenue point is especially important. Investors often treat recurring services as higher-quality growth than product shipments because services can be stickier, easier to model, and more durable across spending cycles. Fortinet’s guidance separated service revenue from total revenue, which gave the market a clear way to judge that mix. A strong total number was not enough if the recurring part of the story looked less robust than hoped.
Reuters also said BofA Global Research downgraded the stock to neutral, and that at least 15 brokerages cut price targets. The firm’s analysts tied some of the debate to component lead times and normalized inventory advantages. That is analyst interpretation, not a Fortinet admission of operational failure, but it helps explain why investors looked past the headline growth figures and asked how repeatable those figures are.
There was also a wider software-market backdrop. ServiceNow fell after its own results, feeding a broader narrative that AI could unsettle established vendors across enterprise technology. That matters as sentiment, but it is not evidence that Fortinet lost customers to AI-native rivals or that AI has already displaced its products. What it does show is that public markets are demanding more proof that AI will deepen differentiation instead of merely raising the cost of keeping up.
What security buyers should take from the disconnect
The practical lesson for buyers is to separate financial signal from security decision.
A 26% one-day stock drop does not show that Fortinet’s controls are unsafe, that customers are leaving, or that the company’s AI-security push has failed. Just as important, strong billings and free cash flow do not prove that every customer is getting better protection, faster investigations, or lower operating cost. Public disclosures still leave open several questions buyers actually care about: customer retention, new-versus-renewal mix, net expansion, AI-security revenue, and whether the newest AI-focused services reduce exposure or false positives in real environments.
That is why procurement and security teams should use moments like this to get more specific in vendor reviews. Ask how much of a proposed deployment depends on appliances versus cloud-delivered controls. Ask for support and update commitments, rollback paths, and realistic deployment latency. Ask how AI inspection works, what telemetry the customer keeps, and how the product integrates with identity, cloud, and incident-response tooling already in place. Ask whether component availability could affect delivery timing. And ask for the all-in operating cost, not just the first contract price.
Fortinet can still benefit if AI expands the threat surface faster than it erodes software differentiation. Enterprises still need firewalls, secure access, cloud controls, and ways to monitor AI use. But the burden of proof is getting heavier. Buyers should want measurable efficacy and manageable operations. Investors, meanwhile, want durable recurring growth and confidence that the AI era strengthens the business model rather than complicating it.
Fortinet’s quarter showed that cyber demand remains real. The stock reaction showed that demand alone is no longer enough to satisfy the market’s harsher scorecard. For technology leaders, that is a useful distinction: buy on protection, integration, service quality, and total cost—not on whether Wall Street applauded the quarter the next morning.
By
By
By
By
By
By
By
By
