NOW company analysis from a product-offering perspective

TLDR: ServiceNows growth likely survives AI commoditization threats because enterprises keep expanding their contracts, not leaving, and $28B in forward contracted revenue suggests enterprises are betting that NOW remains their execution layer

ServiceNow sells two main things. A system of record and a system of workflow. A system of record is where a company stores operational data like employee requests, IT tickets, approvals, and internal service history. Its the source of truth for enterprises A system of workflow is what tells that source of truth what to do with it. If someone submits a request, the workflow routes it, assigns it, escalates it, and tracks it to completion. These are predefined rules that companies configure over time and NOW makes it super easy to configure The key point is how the ServiceNow product actually delights customers

The system of record is necessary, but it is not why they charge premiums because a bunch of existing tools can store data (e.g DynamoDB) at much cheaper prices The system of workflow is what drives revenue and expansion because it provides the customizability for users based on their own personal enterprise needs. It is harder to replace, deeply embedded, and allows NOW to charge a premium. From their investor presentation, a 2010 customer cohort grew their annual contract value by 295%, annually! ($100->$4500) When someone submits an IT ticket, requests time off, or needs an approval, the platform actively routes it, applies business logic (e.g if $500 ask send to manager –> if $1000 IT request, send to Director), escalates when necessary, and closes the loop.

But AI introduces the real tension here.

AI can take a plain language request and determine next steps itself, routing tasks and coordinating across systems without relying on predefined workflows. That directly overlaps with what NOW has historically provided, and the legitimate concern is people may not need to design workflows in advance. In other words, ServiceNow core value is optimizing structured workflows that a user can predefine but AI reduces the need for structure in the first place. The enterprise engineer could provide the mappings to key databases, describe their functionality, upload the API keys and code and provide specific instructions on the outcome desired. LLM would spin it up in minutes but there is nuance

Replicating the function of a workflow is more of a prototype. For serious enterprises where money (or lawsuits) is on the line they need to ensure the workflow uses trusted data, has auditable actions, compliance guardrails, and deep integrations into existing systems. Plus whatever the enterprise engineer whipped up needs to resolve a long list of edge cases, exception handling and basic scenarios before putting it into production. Some enterprises will achieve that. And for the enterprises that do, that is the fundamental bear case. The company built its own replica and can cut ties with NOW entirely.

The bull case, which is where the big push from ServiceNOW is coming from via Agent Fabric, is that agentic AI actually expands their product value. Agent Fabric, is ServiceNows cross-platform orchestration layer that lets its AI agents coordinate with agents from Salesforce, Workday, and other vendors.

Basically, NOW ensures that even in the age of AI, they still provide a service that's even more easier to use More agents across more departments means more orchestration, more governance, and more execution running through the platform. This really increases product stickiness because no one wants to touch something that works so well, across so many vendors and embedded so deeply.

Product stickiness can be seen by the uniform 98% renewal rate every quarter since Q4'24.

So, if NOW can become both the infrastructure layer (system of record), and the orchestration layer across departments and vendors, then effectively this company is able to survive and prosper in the age of LLMs. They're going to get paid no matter how many agents there are, because they'll keep making it easier for you to work. RPO (forward contracted revenue) sits at $28B growing at 26% YoY, which means enterprises are already signing longer and larger commitments.

So it doesn't look like that enterprise engineer is ready to push his workflow to prod just yet!