Is the future of enterprise software less about owning the process and more about being a really good search bar? ServiceNow’s (NOW) stock has plummeted nearly 50% since last summer, a casualty of the AI boom, but the narrative that AI is simply disrupting the company misses the mark. The real story here isn't that AI is a threat to ServiceNow – it’s that AI is forcing ServiceNow to fundamentally redefine its value proposition, and that shift is inherently less profitable. For years, ServiceNow sold businesses the promise of control, of a single, definitive record of how things should work. Now, it’s trying to sell them… a smarter way to ask questions about that record.
ServiceNow built its empire on the Configuration Management Database (CMDB), a frankly unglamorous but essential piece of corporate infrastructure. Think of it as the master blueprint for a company, detailing every asset, every connection, every workflow. Replacing that blueprint isn’t a weekend project; it’s a multi-year, multi-million dollar undertaking. That’s why customer renewal rates are a staggering 98%. But the very tasks ServiceNow automates – routing tickets, coordinating approvals – are precisely the kind of repetitive, rule-based work that generative AI excels at. The irony is brutal: the tool that threatens to replicate ServiceNow’s core functionality is the same technology it now needs to sell.
See the original The Motley Fool story for the full account.
The company’s response is Now Assist, its generative AI offering, which already boasts $600 million in annual contract value and aims for $1 billion by 2026. Sounds promising, right? Except accessing Now Assist requires customers to upgrade to a pricier “Pro Plus” tier and pay usage-based fees. This is where things get tricky. Suddenly, ServiceNow isn’t selling a fixed cost for operational control; it’s selling access to computing power, a commodity where it doesn’t have a natural advantage. Some IT directors are already exploring building their own AI layers on top of existing systems, effectively bypassing ServiceNow altogether. It’s a classic innovator’s dilemma, but with a particularly sharp edge.
This isn’t just about margins, though they are certainly feeling the squeeze. Gross margins dipped 1.5 percentage points last year, a direct result of the escalating compute costs associated with AI. But the deeper issue is the changing physics of the business. Traditional software scales beautifully – once it’s built, distributing it to another user costs next to nothing. AI, however, burns processing power with every query. The more successful Now Assist becomes, the more expensive it is to deliver. Despite this pressure, ServiceNow has managed to expand free cash flow margins to over 34%, fueled by a robust $11 billion commitment to acquisitions focused on AI and security. But acquisitions are a bandage, not a cure.
The core business, while durable, is maturing. Roughly 85% of large companies are already ServiceNow customers. The low-hanging fruit is gone. The company is betting that AI will unlock a new wave of growth, but it’s facing stiff competition from giants like Microsoft, already deeply embedded in the same corporate ecosystems. If employees start defaulting to typing requests into Microsoft Teams and letting an AI agent handle them, the value of ServiceNow’s interface – the very thing it spent years building – diminishes significantly. The market is reacting to this reality, recognizing that ServiceNow’s moat isn’t as impenetrable as it once seemed.
The stock’s decline isn’t simply a reflection of AI hype; it’s a rational response to a fundamental shift in the power dynamics of enterprise software. ServiceNow is transitioning from being the owner of the operational blueprint to being a provider of intelligent access to that blueprint. The question now isn’t whether ServiceNow can compete in the age of AI, but whether it can convince customers to pay a premium for that access when alternatives are rapidly emerging. Watch closely for a surge in companies attempting to build their own internal AI layers on top of existing ServiceNow deployments – that will be the first sign that the blueprint is losing its value.
