A surge of enterprise AI projects now hinges on whether autonomous software agents can interoperate across clouds, data planes, and governance stacks without turning operations into a tangle of brittle integrations or opaque decision chains that risk compliance penalties and service outages. That risk framed a notable shift: ServiceNow and Google Cloud extended their alliance to make agents talk, reason, and act under a single set of controls while still running where data and workloads live. The design connected ServiceNow’s AI Control Tower and Workflow Data Fabric with Gemini Enterprise and BigQuery, aiming for an evidence-based loop from signal detection through diagnosis and resolution. It also tried to balance speed with oversight, leaning on open protocols—such as the Model Context Protocol (MCP)—to avoid the slow drag of walled gardens.
Interoperable Agents: Architecture, Governance, and Real Workloads
At the core, the integration gave enterprises a shared operating picture for agentic AI. ServiceNow AI Control Tower tied into the Gemini Enterprise Agent Platform so policies, lineage, and performance could be enforced in one registry, not scattered across point tools. Workflow Data Fabric served as the bridge between operational records and model context, while BigQuery supplied analytical depth from streaming telemetry to historical joins. This architecture turned a common pain point—lack of consistent context—into a design goal. Agents could exchange state, escalate to humans when thresholds tripped, and log decisions for audits. By privileging MCP and similar open protocols, the partnership reduced the lock-in friction that often stalls cross-platform automation, especially in regulated environments that require reproducible outcomes.
That approach moved quickly from theory to targeted use cases. In telecom, network agents detected jitter spikes or packet loss, used telemetry to infer root causes, then pushed prescriptive actions into ServiceNow workflows that queued field ops or software rollbacks by priority and customer impact. In retail, store and warehouse equipment events landed in BigQuery, where models flagged anomalies that triggered triage, stock holds, and technician dispatch through ServiceNow, keeping returns low and shelf availability stable. For IT operations, cross-cloud agents correlated anomalies across Kubernetes clusters, serverless endpoints, and managed databases, then coordinated remediation to avoid thrashing. The emphasis on one governance layer meant that, regardless of where a model ran, the same policies, approval gates, and telemetry captured the narrative from first signal to final fix.
Financial Momentum: Guidance, Deals, and Strategic Expansion
Business results reinforced the architectural bet. Subscription revenue reached $3.671 billion in the first quarter, up 19% year over year on a constant currency basis, lifting full-year guidance to a range of $15.735–$15.775 billion. Remaining performance obligations rose to $27.7 billion, with current RPO at $12.64 billion, topping internal expectations. Profitability kept pace: a 32% non-GAAP operating margin and diluted non-GAAP EPS of $0.97, while management targeted a 31.5% margin and about a 35% free cash flow margin for the year. Large enterprises leaned in: sixteen deals exceeded $5 million in net new ACV, roughly 80% growth, and 630 customers now spent more than $5 million in ACV, up about 22%. Now Assist accelerated, with customers over $1 million in ACV growing more than 130%, a signal that agentic use cases were expanding beyond pilots.
Building on this foundation, the company widened its footprint through product adjacencies and ecosystem moves. Sales CRM net new ACV grew more than fivefold as deal counts climbed over 80%, suggesting cross-sell traction from workflow incumbency. Moveworks, in its first full quarter post-acquisition, outperformed internal targets and closed more $1 million-plus deals than the entire prior year, validating chatbot-to-agent upgrades in service desks. Security deepened as the Armis acquisition closed early, aligning asset intelligence with incident workflows. Capital allocation stayed active: about 20.1 million shares were repurchased, nearly double the prior year’s total, with $4.2 billion still authorized. Customer outcomes added credibility. TridentCare automated 96% of scheduling across 5.4 million annual patient visits and cut wait times by 57%. Recognition followed, with Google Cloud naming ServiceNow a Partner of the Year in four categories, including Global Business Applications and Agentic AI Innovation.
What Comes Next: Operating Standards for Agentic AI
The partnership’s thesis pointed to a broader market direction: end-to-end automation must connect signals, diagnosis, decisioning, and resolution with uniform governance, or scale will stall under compliance risk and integration cost. For CIOs and heads of operations, the next moves were clear: standardize on interoperable protocols such as MCP to reduce custom glue code; instrument a central registry that tracks agent identity, permissions, model prompts, and audit trails; and anchor high-impact pilots where time-to-value is provable. Telecom fault management, retail asset uptime, and cross-cloud AIOps each offered measurable outcomes—fewer truck rolls, higher shelf availability, faster mean time to resolve—that could be documented in quarterly reviews and used to expand scope. Building reference runbooks that spanned Gemini Enterprise, BigQuery, and ServiceNow workflows also hardened institutional knowledge.
Governance matured in parallel. Teams established policy tiers for low-, medium-, and high-risk automations, mapped human-in-the-loop checkpoints to error budgets, and defined rollback strategies that agents could invoke without human delay. Data stewards synchronized BigQuery schemas with Workflow Data Fabric to keep context current, while security leads bound Armis-derived asset intelligence to incident priorities. Vendors were evaluated on their willingness to support open interfaces, not just headline features, to avoid de facto walled gardens. With financial headroom and visible customer impact, enterprises were positioned to expand agentic coverage from core operations into CRM and security. The practical takeaway had been consistent: interoperability plus disciplined oversight produced durable gains, while closed stacks and ad hoc guardrails raised costs and slowed execution.