ServiceNow Doesn't Sell Software. It Sells a Backlog Most Companies Would Kill For.

A help-desk ticket is the least glamorous object in the modern corporation. Someone's laptop won't connect; they file a request; it sits in a queue; eventually it closes. That is the unpromising thing ServiceNow started with - IT support tickets - and it is the reason most people still file the company under 'enterprise plumbing.' But look at what the plumbing throws off. At the end of 2024, ServiceNow reported $22.3 billion in total remaining performance obligations¹ - revenue customers have already signed for and not yet been billed. That is not a sales pipeline. It is a backlog. The company has, in writing, roughly two years of subscription revenue already contracted and waiting to be recognized.

The official story is that ServiceNow is an IT-ticketing company that got lucky and expanded into other departments. Both halves of that are wrong. It was never about tickets, and the expansion was never luck. ServiceNow is a workflow platform that uses one department's pain to buy its way into all the others - and the money machine is the contract, not the code.

The product is the backlog, not the software

Start with what flows in the door. ServiceNow sells subscriptions, billed annually, to a cloud platform that runs on infrastructure built once and resold endlessly. Q4 2024 subscription revenue was $2.866 billion, up 21% year over year¹ - and that growth had been steadier still earlier in the year, with quarters running 23% to 25%.³ But the income statement is the boring part. The interesting part is the contract that sits behind it. When a customer signs a multi-year deal, most of that revenue can't be recognized yet, so it piles up as RPO - a $22.3 billion mountain of money already promised.¹ Of that, $10.27 billion was current RPO, due within twelve months.¹ A normal software company chases next year's revenue. ServiceNow already has most of it in the bank, in the form of signatures.

Now the magic of the model. A subscription to a platform you've already built costs almost nothing to keep serving - one more customer logging into the same cloud is close to pure margin. And these customers do not leave. In its U.S. Federal business, ServiceNow reported renewal rates near 99%.³ That combination - high-margin recurring revenue that compounds because almost no one churns - is why the cash piles up rather than getting consumed. Free cash flow hit $3.415 billion in 2024, up about 26% from the year before, and then $4.576 billion in 2025, up roughly a third again.⁴ The backlog pre-funds the growth; the growth refills the backlog. The flywheel runs on the fact that nobody unplugs.

Why IT was the door, not the building

Here is where the popular story goes most wrong. The narrative is that ServiceNow started with IT ticketing and then, almost by happy accident, discovered it could also do HR and customer service and security. The accurate version is the opposite: Fred Luddy designed the platform from inception as a general workflow engine.¹¹ IT service management was the chosen beachhead - the first room to enter - but the platform was built to be adaptable far beyond IT from the start.¹¹ The reason IT was the right first room is subtle. IT managers sit at the center of every other department's processes; win them, and you've won a credible reference inside the building who can vouch for you to HR, to facilities, to legal. Sequoia, an early backer, describes exactly this: IT managers became the beachhead for organic, enterprise-wide expansion.⁸ You don't sell a workflow platform department by department by cold-calling. You sell one workflow, embed it, and let the org pull you in.

You can watch this work in the customer data. By the end of 2024, nearly 500 customers were spending more than $5 million a year with ServiceNow, a count growing 21% year over year.¹ These are not companies that bought a ticketing tool. These are companies that started with one workflow and kept adding more until ServiceNow ran a meaningful slice of how the whole enterprise operates. And once that many of your processes live on one platform, ripping it out becomes its own multi-year project nobody wants to sponsor. The 99% renewal isn't loyalty. It's gravity.

That phrase is worth pausing on, because it's the company's own marketing, and the load-bearing word is 'organically.' ServiceNow didn't buy its way to $7 billion through acquisitions stapled together. It grew the platform from inside its own customers - the same expansion mechanic, compounded over a decade. The founding itself fits the pattern of unglamorous beginnings: the product began life as Glidesoft in 2003, with Luddy as the only employee until mid-2005, when a modest $2.5 million round from JMI Equity finally funded five additional hires.⁵ Luddy came to it after his roughly $35 million in stock at Peregrine Systems, where he had been chief technology officer, was wiped out in that company's collapse.⁶ He didn't build a ticketing tool. He built a workflow engine and walked it in through the IT door.

Isn't all of this already priced in?

The honest objection is that none of this is a secret. A near-100% renewal rate, a $22.3 billion backlog, a platform that expands itself - the market sees all of it, and has bid the stock accordingly. The bull case isn't hidden; it's consensus. So the real question isn't whether the moat is wide. It plainly is. The question is whether the growth rate can survive the law of large numbers. A company adding 20%-plus to revenue is one thing at $3 billion and quite another as it crosses past $11 billion in annual revenue - the same percentage now demands a vastly larger pile of net-new contracts every single year. The early days, when ServiceNow beat its first board-meeting plan by 70%,¹⁰ are not the math it faces now. The stress-test is whether new bets - agentic AI, and acquisitions like Armis, expected to add about 125 basis points to 2026 subscription growth while costing roughly 200 basis points of free-cash-flow margin⁹ - can keep the backlog compounding at that pace, or whether scale finally bends the curve down.

The fair answer is that the early evidence still points up. Current RPO reached $12.64 billion in the first quarter of 2026, up about 22.5% year over year, and the count of customers spending more than $5 million a year had climbed to 630.⁹ The expansion engine is still expanding. But notice the cost of the acquisitions ServiceNow is now reaching for: a deal that adds revenue growth while subtracting cash-flow margin is the trade a company makes when pure organic expansion starts to need help. That isn't a crack in the moat. It's the sound of a $11-billion business working harder to move a needle that used to move on its own.

ServiceNow makes its money the way a utility does, except it gets to keep adding new services to the same meter. The unglamorous ticket was never the business. It was the entry point - the cheapest possible way to get one workflow inside a building so the rest could follow, room by room, contract by contract, until the company was running enough of the enterprise that leaving stopped being a decision anyone wanted to make. The backlog is the proof. Two years of revenue, already signed, before the year even starts. The hard part now isn't winning the building. It's that ServiceNow has already won so many of them that the next one barely moves the number - and that, not competition, is the only real threat to the machine.