ServiceNow shares plunge 24% after earnings revive AI disruption fears

ServiceNow shares sank 24% on Thursday, heading for their worst day ever, after Q3 adjusted EPS of $1.00 missed the $1.02 forecast; revenue rose more than 20% to $4.11 billion, roughly matching estimates. The company slightly raised its full-year 2026 subscription revenue outlook to $15.78 billion-$15.79 billion, and net new AI average contract value jumped 65%. Investors remained concerned that AI investments may take longer to produce results and that AI could disrupt traditional software business models.

Analysis

The selloff looks more like a credibility and duration reset than evidence of an immediate revenue cliff: investors were paying for reliable execution and an earlier AI monetization curve, so a small earnings miss can drive disproportionate multiple compression even with guidance intact. The key second-order risk is that enterprise AI may shift value from seat-based software toward usage, platforms, and implementation—pressuring SaaS pricing before new AI revenue is large enough to offset it. Conversely, embedded engineering could strengthen retention and expand contracts, but only if deployments convert into recurring subscription dollars without structurally higher delivery costs.

Over the next 1–3 months, watch subscription backlog/current remaining performance obligations, renewal and expansion trends, and evidence that AI contract value is recognized as durable subscription revenue. Also verify forward-deployed engineering costs and whether management changes hiring or margin expectations. Over 6–18 months, the contest is whether AI increases NOW’s workflow control and wallet share or lets customers/custom-built agents bypass paid seats. The raised outlook argues against treating this quarter alone as proof of disruption; it does not resolve the timing or economics. A further guide or backlog downgrade would falsify the “valuation reset, not deterioration” view. A sustained stabilization in estimates and AI-to-subscription conversion would weaken the bear case.

Key Decisions for Investors

• Avoid chasing the initial gap lower: the sharp repricing raises squeeze risk, while the near-term evidence is not enough to establish a broken subscription model.
• For a 1–3 month stance, keep NOW exposure below benchmark weight until the next data point confirms backlog and subscription growth; reassess on estimate revisions, not AI contract-value headlines alone.
• Watch for an eventual relative-value entry—long NOW versus a broad software basket only if backlog/renewals hold and AI deployments convert to recurring revenue without margin deterioration. If estimates fall or guidance is cut, do not treat the prior drawdown as a floor.
• Key falsifiers: successive subscription/backlog deceleration or lower guidance would validate structural concern; improving backlog plus stable delivery costs and recurring AI conversion would support recovery. No precise valuation-based target is justified from the supplied data.