Everything About AI
NextFin News - ServiceNow shares tumbled as much as 17% on Wednesday before paring losses to close 3% lower, after the enterprise software maker's fourth-quarter subscription sales outlook merely met Wall Street's elevated bar instead of blowing past it - reigniting the market's deepest fear that artificial intelligence will disrupt, rather than extend, the company's seat-based licensing model.
The company reported third-quarter revenue of $4.11 billion, up 20.5% year over year and in line with the $4.10 billion consensus, while non-GAAP earnings of $1.00 a share missed estimates of $1.02 by 2.3%. Adjusted operating income of $1.34 billion beat the $1.28 billion expected, implying a 32.6% adjusted margin. Billings rose 23.2% to $3.64 billion. Yet investors focused on the forward-looking metrics: subscription revenue is expected to grow as much as 23% in the current quarter, about in line with analyst estimates, while current remaining performance obligations - a closely watched bookings proxy - are set to expand more than 18%, just shy of the 19% consensus.
The reaction captures the new arithmetic of the AI trade: a beat-and-raise quarter is no longer enough when the market is pricing in a structural transformation of the software business itself.
What the Numbers Actually Said
ServiceNow's third-quarter report was, on its face, a solid execution quarter. Revenue grew 20.5% to $4.11 billion, matching the consensus mark. Adjusted operating income of $1.34 billion came in 4.8% ahead of estimates. Free cash flow margin improved to 20% from 15.9% in the prior quarter. And Chief Executive Bill McDermott struck an emphatically bullish tone, calling the quarter "the inflection point for ServiceNow as the Super Intelligence Control Tower" and pointing to a hybrid business model that had surged to $30 billion in remaining performance obligations, growing 23%.
"This was more than another beat-and-raise quarter for ServiceNow, this was the inflection point for ServiceNow as the Super Intelligence Control Tower. Our customers have validated our strategic relevance as the single harness for enterprise intelligence and security."
The company also reported that its AI annual contract value crossed $1 billion in the second quarter, with agentic deployments increasing ninefold over nine months and deals including five or more AI products growing 5.5 times year over year. Management said it remained on track to beat its $1.5 billion year-end AI ACV target. AI-native SKUs carry price premiums in the range of 20% to 30%, according to research coverage of the company's pricing.
But the guide told a different story at the margin. Subscription revenue growth of up to 23% for the fourth quarter landed squarely in line with expectations rather than above them. And cRPO growth of more than 18% - the metric investors watch most closely for early signs of demand acceleration or deceleration - fell just short of the 19% average projection compiled by analysts. That single percentage point of shortfall was enough to send the stock down as much as 17.2% to $124.46 in the immediate aftermath, before it recovered to close at $150.13, down $4.65, or 3%, on the day.
The deceleration is visible even in the company's own trajectory. In the second quarter, cRPO grew 21% year over year, ending the period at $13.2 billion. The third-quarter guide of more than 18% represents a three-point step-down in the growth rate of the very metric that had been the bull case's strongest pillar. A three-point deceleration in bookings growth is not a rounding error; it is the first quarterly confirmation that demand is not accelerating even as AI hype reached a fever pitch.
Beneath the headline profitability beat, the GAAP picture was less comfortable: GAAP operating margin compressed to 9.4% from 16.8% in the year-ago quarter. That divergence - adjusted strength alongside GAAP compression - is the fingerprint of a company spending heavily on AI infrastructure, compute, and compensation while trying to prove the spending will pay off. It is exactly the trade-off that has made investors skittish across the software sector: the costs of building AI capabilities hit the GAAP income statement today, while the revenue payoff arrives, if at all, over several years.
The combination is what unnerved investors: the company is still growing at roughly 20%, but the market had been pricing it as if AI would accelerate that growth, not merely sustain it. When the forward bookings metric failed to confirm acceleration, the premium evaporated.
Why a Beat-and-Raise Quarter Got Punished
The surface explanation is simple - the guide was merely in line. The deeper mechanism is that ServiceNow's valuation had run ahead of its fundamentals on AI enthusiasm, and the report forced a reckoning. Shares had rallied roughly 43% in the three months leading into the print, trading at about 33 times forward earnings, with the market effectively pre-paying for several years of uninterrupted AI-driven acceleration. The stock had also climbed back only partway from a year-to-date decline that left it down roughly 38% from its peak - a sign that even after the rally, investors were holding the shares on a knife's edge.
At those multiples, "in line" is a failure. A company valued for perpetual acceleration cannot afford a single quarter where the bookings engine merely holds steady. The cRPO miss - however small in absolute terms - was read as the first visible crack in a narrative that had carried the stock: that AI would compound, not cannibalize, ServiceNow's core workflow business.
This is the second-order risk that the market is now pricing, and it extends well beyond ServiceNow. The same capital that poured into AI infrastructure - chips, data centers, cloud capacity - is now being asked to show a return in the form of software productivity gains. If AI agents make knowledge workers more productive, enterprises may need fewer software seats to get the same work done. ServiceNow's traditional licensing model, built on per-seat, per-year subscriptions, is directly exposed to that dynamic. The fear is not that AI will fail; it is that AI will succeed so well that it shrinks the very seat count the software industry charges for.
That is why the market punished a company that, by conventional metrics, executed well. Investors are not evaluating this quarter's revenue. They are underwriting a regime change in the unit economics of enterprise software - and they are discounting the possibility that the seat-based model that produced 20% growth for a decade cannot survive the agentic future intact.
The transmission channel runs through the P&L in both directions. On the cost side, AI raises expenses immediately: higher compute bills, more data-center capacity, premium engineering talent, and the sales effort required to convince customers to adopt new AI products. On the revenue side, the payoff is uncertain and delayed: enterprises must first prove that AI agents deliver measurable productivity before they expand contracts. In the gap between those two timelines, margins compress - and that is precisely what the GAAP print showed.
Cyclical Wobble or Structural Repricing?
The critical question is whether this selloff is a cyclical overreaction to one quarter's guide or the start of a structural re-rating of the software sector. The evidence points to structural.
A cyclical decline would be tied to a transient cause - a one-time federal contract timing shift, a macro slowdown in IT spending, or a temporary digestion period after a large acquisition. ServiceNow has faced those before and mean-reverted. This episode is different. The selloff occurred despite revenue in line with estimates, adjusted operating income ahead of estimates, raised full-year guidance, and a CEO declaring an "inflection point." A cyclical story does not sell off on good news. A structural story does.
Three pieces of evidence support the structural read. First, the valuation compression is sector-wide, not company-specific: enterprise software peers have faced the same "death of SaaS" scrutiny as investors question whether AI will reduce software headcount and, by extension, software spend. Second, the trigger is technological, not financial - agentic AI changes the unit of consumption from seats to outcomes, and no amount of quarterly execution can reverse a change in the unit of consumption. Third, the market's reaction function has shifted: before the AI cycle, a beat-and-raise quarter was rewarded; now, it is punished unless it also confirms acceleration in the very metric - cRPO - that would prove AI is additive rather than substitutive.
History offers a cautionary parallel. During the shift from on-premise software to cloud subscriptions in the early 2010s, incumbent vendors that failed to reposition their licensing models saw their multiples compress permanently, even when their revenue kept growing. The market did not wait for the revenue to roll over; it repriced the model first and asked questions later. ServiceNow's management is arguing that AI is the opposite of that disruption - that the Control Tower becomes more valuable, not less, as AI proliferates. The market, so far, is not convinced.
The counter-thesis is serious and deserves weight. ServiceNow is not a passive victim of AI; it is one of the clearest monetizers in enterprise software. Its AI Control Tower and Now Assist products carry price premiums of 20% to 30%, AI ACV crossed $1 billion with deployments up ninefold, and deals with five or more AI products grew 5.5 times. Management argues that AI increases the complexity enterprises must manage - more agents, more models, more governance, more security - making the Control Tower more essential, not less. Under this view, Wednesday's selloff was an emotional overreaction to a one-percentage-point cRPO shortfall in an otherwise strong quarter, and the stock's 43% three-month rally had simply left no room for any guide that wasn't perfect.
That argument is coherent, but it rests on a premise the market has not yet accepted: that AI revenue will grow fast enough to offset any seat-based erosion. The falsifying signal is concrete. If ServiceNow's fourth-quarter subscription growth lands at the high end of its 23% guide while cRPO growth re-accelerates above 19% and AI ACV continues to compound toward and beyond the $1.5 billion year-end target, the disruption thesis is wrong - AI is additive, and the selloff was a buying opportunity. If instead cRPO growth stalls in the mid-teens while AI revenue remains a single-digit percentage of the total, the structural repricing has further to run.
What Comes Next
In the short term, the stock will trade on sentiment and positioning. A 17% intraday plunge followed by a partial recovery signals capitulation among momentum holders but also the first wave of value interest at lower multiples. Volatility will remain elevated into the next print, and the options market will likely price in continued two-way risk around each monthly AI announcement and each peer earnings report.
Over the medium term, the fundamental question is the composition of growth. Investors will watch whether AI ACV scales from $1 billion toward the $1.5 billion year-end target and beyond, and whether the mix of revenue keeps shifting toward products least exposed to seat-count erosion. The $30 billion RPO base, growing 23%, provides a cushion; the question is whether that growth rate holds or decelerates. The fourth-quarter guide - subscription growth of up to 23% and cRPO growth of more than 18% - sets a low bar for the next report, which is the first place a recovery could begin.
Over the long term, the outcome turns on a single structural question: does agentic AI expand or contract the total addressable market for workflow software? If AI makes every employee a power user who needs more automation, ServiceNow's TAM expands and the current multiple looks cheap. If AI lets companies do the same work with fewer licensed seats, the multiple compresses toward the broader software average and the stock has further to fall.
Base case: ServiceNow grows subscription revenue in the high teens to low twenties for the next several quarters, AI revenue scales but remains a minority of the total, and the stock trades in a wide range as the market waits for proof that AI is additive. Upside case: cRPO re-accelerates above 19%, AI ACV compounds past $1.5 billion, and non-seat-based revenue becomes the majority of new bookings - the disruption thesis breaks and the premium multiple returns. Downside case: cRPO growth decelerates into the mid-teens while AI revenue fails to offset seat-based softness - the sector re-rates lower and ServiceNow follows.
The market spent the last year paying ServiceNow for the promise that AI would accelerate its growth. Wednesday's verdict was harsh but internally consistent: until the bookings data proves AI is additive, the market will price the company as if AI is a threat. The next quarter will decide which it is.
Data as of the October 28, 2026 market close.
Explore more exclusive insights at nextfin.ai.
