ServiceNow Grew Subscriptions 21% — the 18.5% Forward Guide Is What the Market Actually Traded

ServiceNow's third-quarter 2026 results landed with a familiar split verdict: the reported business expanded briskly, and the stock still came under heavy pressure on Thursday. Subscription revenue climbed 21% year over year to $3,988 million, and adjusted diluted earnings per share came in at $1.00. Media reports indicated that adjusted EPS also fell short of consensus estimates, so the market's attention shifted almost entirely from what had already been booked to what management expects next.

That forward view is where the real friction sits. For the fourth quarter of 2026, ServiceNow guided subscription revenue to a range of $4,245 million to $4,255 million. Growth in current remaining performance obligations — the cRPO metric covering contracted work not yet delivered — is projected at 18.5%. According to Reuters, that guidance trailed market expectations, and the resulting selloff in software names, which also followed disappointing results from Fortinet, reflected anxiety about how artificial intelligence might disrupt established software vendors.

The Guide, Not the Partner List, Is the Test

Anyone trying to judge ServiceNow's future should weight that cRPO outlook more heavily than the roster of new collaborators. The distinction matters: a cautious forecast has sharpened fears of AI-driven displacement, but on its own it does not prove that customers are leaving ServiceNow because of artificial intelligence. There is a wide gap between a skeptical market narrative and a documented change in the business.

Backward-looking growth cannot answer the question of future demand — which is precisely why neither a purely bullish nor a purely bearish reading holds up. Strong quarterly numbers do not automatically neutralize a conservatively framed outlook, and a disappointing guide does not render prior growth meaningless. The forecast has intensified concerns about possible AI cannibalization; it has not demonstrated them.

Two Analysts, One Direction

The analyst response roughly a week ago captured that tension. Morgan Stanley cut its price target to $160 from $180 while keeping an "Overweight" rating, citing slower cRPO growth and the fourth-quarter outlook. Stifel, by contrast, reaffirmed both its equally high price target and its buy recommendation.

The verdict was therefore not uniform — one house trimmed its target, the other saw no reason to adjust. Yet both maintained their positive investment ratings, which separates criticism of the near-term growth pace from a fundamentally negative view of the company.

Customer Deals Supply the Operating Counterweight

Concrete deployments offer the other side of the ledger. On October 28, ServiceNow announced a five-year agreement with HPE, which selected the ServiceNow AI Platform to support its Cloud & AI business — spanning customer service, field service and AI governance, among other areas. Such agreements demonstrate real-world applications of the platform, though they do not by themselves answer whether ServiceNow can meet elevated growth expectations next quarter.

A second example arrived on October 22. At customer Thrive, the deployed CRM solution with AI agents reached a 73% first-contact resolution rate, according to the company, while time to resolution fell 63%.

Partnerships Are a Strategy, Not a Result

ServiceNow is meeting the AI shift head-on through alliances. Its expanded collaboration with OpenAI, announced roughly a week ago, includes a ServiceNow plugin for ChatGPT, alongside planned integrations around Codex, security and risk solutions, and voice-driven governance. The Control Tower Alliance, unveiled around the same time, counts Anthropic, AWS, Cisco, NVIDIA and OpenAI among its founding members and aims to support enterprise AI adoption.

Those are strategically coherent moves — ServiceNow is working to connect its offerings with AI applications. But no reliable statement about incremental revenue follows from them. The wording deserves close attention: existing components of a partnership are not the same as planned integrations, and a plan is not a closed product deal. The prominent names make the initiatives noteworthy; they do not substitute for contract growth or its later conversion into revenue.

The Yardstick That Matters

On balance, a wait-and-see, critical stance looks more persuasive than a blanket call on who wins or loses in the AI transition. ServiceNow is posting growth and pursuing tangible collaborations, while the order outlook deserves more weight than the sheer number of partners.

The opportunity lies in linking ServiceNow's own offerings with those partners' applications. The risk lies in mistaking strategic announcements for commercial success. Contract and revenue growth remain the measure — that is what the AI strategy must be judged against, not how impressive its partner list sounds. Measurable customer outcomes now sit opposite a cautiously received outlook, and the decisive question is whether those deployments and agreements translate into growth that clears the market's bar. New partnerships alone will not relieve that valuation pressure.