NOW

ServiceNow has been hit alongside software generally, because investors increasingly view AI agents as a threat to traditional software. The concern is coherent: if an agent can perform the work a seat-based subscription used to mediate, then seat counts and pricing power both come under pressure.

The weakness looks more like a valuation discount than a deterioration in the opportunity. That distinction is the entire position, and it is also the risk: a re-rating driven by a structural fear does not reverse until results disprove the fear. Agents could also compress seat-based pricing outright, which would change the economics rather than just the multiple.

ServiceNow could instead become one of the major beneficiaries of agentic AI. Workflows and enterprise automation become more valuable when software agents can actually execute them, not less. An agent needs a system of record, a permission model and an auditable process to act inside. That is what the platform already is.

The measurable version of this argument is subscription revenue growth and margin, since a platform genuinely capturing agent workloads should show expansion rather than seat erosion.