Okta's growth halves to 11% while the GAAP-to-adjusted gap swallows half its profit

---

There is a particular kind of corporate adolescence that no investor deck ever names: the moment a hypergrowth company stops growing fast and has to decide what story to tell instead. Okta is living that moment in public. On May 28, 2026, the identity-and-access vendor reported its first quarter of fiscal 2027 — the three months ended April 2026 — and the headline number was the one nobody at the company wanted to be the headline. Revenue grew eleven percent year over year, to $765 million. Subscription revenue, the part that is supposed to be the durable engine, also grew eleven percent. For a company whose entire valuation history was built on the premise that identity is the control plane of the cloud, and that the control plane compounds, eleven percent is not a number. It is a verdict.

To understand how far the slope has tilted, you have to remember where Okta came from. In its hypergrowth years the company routinely posted revenue growth north of fifty percent, and the market paid for it the way the market always pays for software that seems to grow without limit — with a multiple that assumed the line would bend up, not down. Today the line bends down. Management's own guidance for full-year fiscal 2027 calls for revenue of $3.185 billion to $3.205 billion, a growth rate of nine to ten percent. That is the slowest annual growth Okta has guided to as a public company. The deceleration is not a single soft quarter to be dismissed; it is the company's own forward map, signed by its own finance team. The question this article asks is simple and unsentimental: is Okta a secular grower having a bad cycle, or a maturing utility being priced, in places, as if the bad cycle were temporary?

The number management would rather you watch instead

Every decelerating software company eventually develops a favorite metric, and it is rarely revenue. Okta's preferred lens is remaining performance obligations — the contracted backlog of subscription dollars not yet recognized. And here the company has a genuinely better story to tell: total RPO surged past $4 billion, reported at $4.084 billion, growing in the high teens to low twenties on a percentage basis depending on the quarter. That is a real number and a real strength, and we will give it its due later. But notice the sleight of attention. Total RPO includes contracts stretching years into the future, and long-dated backlog is precisely the metric that flatters when customers sign longer deals rather than bigger ones. The cleaner read on near-term demand is current RPO — the backlog expected to convert to revenue over the next twelve months. And current RPO grew twelve percent, to $2.499 billion. Twelve percent. The figure that actually predicts next year's revenue is growing in lockstep with revenue itself, which is to say, it offers no acceleration to lean on.

This is the denominator illusion in miniature. When a company wants its growth to look better than its income statement, it points you at the longest-dated, most-aggregated version of its backlog and lets the big absolute number — four billion! — do the emotional work. The number is true. The implication, that demand is reaccelerating, is not supported by the twelve-percent current-RPO print sitting right next to it. A backlog that grows because contracts lengthen is a different animal from a backlog that grows because more customers are buying more product faster. Okta's investors are being invited, gently, to conflate the two.

The hundred-million-dollar wedge between two kinds of profit

Now to the adjustment. Okta reported GAAP net income of $74 million in the quarter, up from $62 million a year earlier — a genuine, real, cash-relevant improvement, and a milestone for a company that spent most of its public life unprofitable on a GAAP basis. But the number the company leads with, and the number the Street models, is non-GAAP net income of $168 million. The gap between $74 million and $168 million is roughly $94 million in a single quarter, and the largest single line bridging the two is stock-based compensation — the practice of paying employees in shares the company prints rather than cash it earns.

There is a respectable accounting argument that stock comp is a non-cash expense and that free cash flow is the truer measure. Okta will happily make it; free cash flow was a robust $271 million in the quarter. But the argument has a hole the size of the share count. Stock-based compensation is non-cash only in the narrow sense that no dollar leaves the building today. Its cost is borne by existing shareholders through dilution — every share granted is a sliver of the company handed to an employee instead of an owner. To exclude it from "profit" while it quietly expands the denominator of every per-share figure is to count the benefit of cheap labor and ignore the bill. When a company's non-GAAP profit is more than double its GAAP profit, the quality of that earnings is, at minimum, a question worth asking out loud. Okta is more profitable than it used to be. It is not yet as profitable as its headline numbers, repeated often enough, can make it feel.

The bundle problem nobody on the call wants to name

Okta's commercial existence rests on a single elegant proposition: identity should be neutral. A company runs Microsoft and Google and Salesforce and AWS and a hundred SaaS tools, and Okta sits in the middle, vendor-agnostic, federating logins across all of them. For years that neutrality was a moat. The trouble is that one of those vendors — Microsoft — gives away a competing identity layer, Entra ID, bundled into the Microsoft 365 and Azure subscriptions that most enterprises already pay for. When the identity product is "free" in the sense that it arrives inside a license you have already bought, the procurement math turns hostile fast. A CFO does not need Entra to be better than Okta. The CFO needs Entra to be good enough, and already paid for.

This is the moat-versus-loophole problem at the heart of the bear case. Okta's neutrality is a real advantage for heterogeneous environments — but the gravitational pull of the Microsoft estate is relentless, and every quarter a Microsoft shop decides "good enough and included" beats "better and a separate invoice" is a quarter Okta has to win net-new logos elsewhere just to stand still. The eleven-percent growth rate is the sound of that treadmill. Management rarely names Entra on earnings calls without immediately pivoting to Okta's independence and breadth, and the breadth is genuine. But you cannot bundle your way out of a bundling problem, and Okta is structurally on the wrong side of the largest software distribution machine ever built.

Demonstration versus deployment: the new-product carousel

Confronted with a maturing core, Okta has done what every maturing software company does — it has launched adjacencies. Identity Governance. Privileged Access. Device Access. Fine Grained Authorization. Identity Security Posture Management. Identity Threat Protection with Okta AI. It is an impressive catalog, and management cites it as evidence that the platform is expanding, that the addressable market is widening, that growth will reaccelerate as these new products land. The catalog is the bull case rendered as a product roadmap.

Here is the forensic distinction that matters: demonstration is not deployment. A product can exist, demo beautifully, win a Gartner mention, and appear in every customer pitch while contributing a rounding error to revenue. The proof that new products are driving the business is not their existence; it is acceleration in the metrics that aggregate them — and Okta's current RPO is growing twelve percent and its revenue eleven. If governance and privileged access and the rest were converting from demo to deployment at scale, you would see it in those numbers as upward pressure, not as a flat line dressed in a longer catalog. In privileged access specifically, Okta is selling into a market owned by specialists: CyberArk, the category's reference implementation, was acquired by Palo Alto Networks in a roughly $25 billion deal that closed February 11, 2026 — which means Okta's privileged-access ambitions now collide not with a standalone vendor but with one of the best-distributed security platforms in the industry. The new products are real. Whether they are deployments or demonstrations is, for now, an open question the income statement has not yet answered in Okta's favor.

Cyclical priced as secular, and the breach that never fully closed

Software bulls have spent two years arguing that the great deceleration across SaaS is cyclical — a hangover from over-provisioning in the zero-rate years, to be cured when budgets thaw. Maybe. But Okta carries a company-specific scar that complicates the cyclical story. In late 2023, attackers compromised Okta's customer support case-management system and, through it, reached files belonging to a meaningful slice of its customers. For an identity-security vendor — a company whose entire product is the promise that it can be trusted to guard the keys — a breach of its own support systems was not just an incident. It was an existential advertisement for the bear case. Trust, once dented, does not re-grow on the company's preferred timetable. Some portion of Okta's deceleration is almost certainly the lingering cost of that episode in renewal conversations and competitive bake-offs, and that portion is not cyclical at all. It is reputational, and it is Okta's alone.

The market's error, if there is one, is treating the entire eleven-percent print as a temporary cyclical trough that will mean-revert toward the old growth rate once IT budgets recover. That is the cyclical-priced-as-secular trap in reverse: assuming a secular reset is a cyclical dip. If a chunk of the slowdown is Entra's structural pull plus a durable trust discount, then "wait for the budget cycle to turn" is not a thesis. It is a hope.

Priced for a recovery the guidance does not promise

Okta's equity carries a market capitalization in the neighborhood of $14 billion, and the analyst community, on balance, still rates it a buy, with a median price target around $100 and a range that runs from the mid-seventies into the $140s. Valuation multiples vary wildly by source and by which earnings figure you trust — trailing GAAP P/E readings cluster in the high-fifties to mid-eighties, forward P/E on non-GAAP earnings in the twenties to thirties. That dispersion is itself the tell: when reasonable observers cannot agree whether a stock trades at twenty times earnings or eighty, they are disagreeing about which "earnings" is real — the GAAP number or the adjusted one — which is exactly the wedge this article has been circling.

Here is the asymmetry. To justify a premium multiple, Okta must reaccelerate. To merely hold its multiple, it must at least stop decelerating. But its own guidance points the other way: nine to ten percent for the full year, slower than the eleven it just printed. A stock priced for a turn, guided toward a continued slide, is the textbook setup for a priced-for-perfection disappointment — not because the company is failing, but because the company is succeeding at being exactly what a mid-single-digit-to-low-double-digit grower is, while a residue of its valuation still remembers what it used to be.

The professional-services tell

There is a small, easily overlooked admission buried in Okta's full-year guidance, and it is the kind of detail forensic readers live for. Management disclosed that fiscal 2027 revenue growth carries an approximately one-percentage-point headwind from a deliberate decision to push professional-services work to partners. On its face this is a margin-optimization move — services are low-margin, so shedding them lifts the blended profile. But read it the other way. A company confident in reaccelerating top-line growth does not usually volunteer to shave a point off the headline. Okta is choosing reported-margin quality over reported-revenue growth, which tells you where management's own confidence sits: in the durability of profitability, not in the reignition of growth. That is a rational choice for a maturing business. It is not the choice of a company that believes the next chapter is hypergrowth redux.

The seat-count ceiling and the AI cross-current

There is a structural ceiling hiding in Okta's business model that the artificial-intelligence narrative both threatens and, in management's telling, rescues. Workforce identity is priced, fundamentally, per human seat — you pay Okta for each employee who needs a login. But the headcount of large enterprises is not growing the way it did, and in pockets it is shrinking as automation absorbs roles. A business whose revenue scales with the number of human employees at its customers is exposed, quietly, to the precise trend its own marketing celebrates: if artificial intelligence makes companies more productive with fewer people, the seat-count denominator that Okta bills against stops expanding. Management's answer is that machine identities — the non-human accounts, agents, and services that proliferate in an AI-saturated stack — represent a vast new population to secure and therefore to charge for. That may well be true, and it is the most intellectually honest version of the bull case. But it is also, today, a promise rather than a line item. The pivot from billing humans to billing machines is a thesis Okta is asserting, not a transition its current revenue growth reflects. Until the machine-identity opportunity shows up as acceleration in current RPO rather than as a slide on the earnings call, it belongs in the demonstration column, alongside the rest of the new-product carousel — a credible future that the present income statement has not yet begun to corroborate.

What the bulls genuinely get right

Now the concession, and it is a real one, because the bear case here is about valuation and trajectory, not about a broken company. Okta is not a fraud, not a melting ice cube, and not without genuine strengths — and a fair forensic account has to say so plainly.

First, the cash generation is the real article. Free cash flow of $271 million in a single quarter, and full-year fiscal 2027 free-cash-flow guidance of $855 million to $885 million, is not the profile of a struggling business. Okta converts revenue to cash at a rate that many faster-growing peers would envy, and that cash funds buybacks, product development, and resilience without diluting the balance sheet. A company throwing off this much cash has time and optionality that a cash-burning grower does not.

Second, the backlog is genuinely large and genuinely sticky. Total RPO past $4 billion, even granting the denominator critique above, represents contracted commitments that do not evaporate in a soft quarter. Identity is mission-critical infrastructure; once a large enterprise standardizes on Okta across thousands of applications, ripping it out is expensive, risky, and slow. The switching costs that make Entra's bundling a slower threat than the bear case implies are the same switching costs that protect Okta's installed base. Neutrality remains a real selling point for genuinely heterogeneous enterprises that refuse to hand Microsoft the identity layer along with everything else.

Third, the margin transformation is real and durable. Non-GAAP operating margins of twenty-six percent, with full-year guidance of twenty-five to twenty-six percent, mark a company that has genuinely learned to operate profitably — a hard transition many software names never complete. And the new-product portfolio, demonstration-versus-deployment caveats notwithstanding, is aimed at exactly the right places: governance and identity-threat protection are where security budgets are actually growing, and Okta has credible products there, not vaporware. If even one or two of those adjacencies cross from demo to deployment at scale, the growth math improves materially. The bull case is not stupid. It is a bet that a profitable, sticky, cash-rich infrastructure company with a widening product surface will eventually convert that surface into reacceleration. That bet could win.

The kicker

What it cannot do is win on the current numbers. The bull case is a forecast; the bear case is the income statement. Eleven-percent growth decelerating toward nine, a GAAP profit less than half the adjusted one, a backlog whose near-term slice grows no faster than revenue, a free identity layer bundled by the largest software company on earth, and a privileged-access ambition now staring down a Palo Alto–owned CyberArk — these are not allegations. They are filings. The question is not whether Okta is a good company. It plainly is. The question is whether a good company growing at single digits, earning half its headline profit in stock it prints, and carrying a reputational scar that an identity vendor can least afford, should still trade with a premium attached to the memory of fifty-percent growth. The market has not yet decided. The guidance already has.

Okta is no longer the rocket; it is the utility that used to be the rocket, and the gap between those two stories is exactly the room left in the stock to fall.