ServiceNow Grew Revenue 20% and the Stock Still Fell. Will AI Agents Replace SaaS?
ServiceNow raised its AI contract-value target to $1.7 billion, yet shares dropped. Meanwhile Serval, a startup that wants to replace it, tripled its valuation in about 10 months.
ServiceNow, a heavyweight in enterprise software, grew third-quarter revenue 20.5% and raised its target for AI-related annual contract value to $1.7 billion. The stock still fell sharply. Over roughly the same period, Serval, an AI startup that says it will replace ServiceNow, tripled its valuation in about 10 months.
If AI agents handle work in place of employees, does existing software lose its place? Or do incumbent platforms that control a company's data and security systems come out stronger?
1. Why did the stock plunge on good results?
ServiceNow reported solid growth in its third-quarter 2026 results on Oct. 28.
| Key metric | Q3 2026 |
|---|---|
| Total revenue | $4.106 billion (+20.5% YoY) |
| Subscription revenue | $3.988 billion (+21% YoY) |
| cRPO | $13.65 billion (+20.5% YoY) |
| Total RPO | $29.7 billion |
| 2026 AI-related ACV target | Raised to $1.7 billion |
cRPO is contracted backlog that will be recognized as revenue over the next 12 months. It is a useful gauge of future revenue visibility. Total RPO covers a longer period.
AI-related ACV is a measure of annual contract value. The $1.7 billion does not mean AI revenue recognized in 2026 accounting.
Even so, the shares fell sharply on Oct. 29. Investors appeared to worry less about the results already reported than about the pace of future subscription revenue and backlog growth.
A growth stock's price reflects expectations for future growth as well as current earnings. A strong quarter and a falling share price can therefore happen together. The drop alone does not prove the SaaS business model is breaking down.
2. SaaS's real strength is not the screen
SaaS (software as a service) is a model in which companies pay a subscription to use software instead of installing and owning it.
The strength of an enterprise platform like ServiceNow comes from three places.
- Switching costs. Moving the workflows and management systems of thousands of employees to another system takes time and money.
- Data and system integration. HR, IT, finance and customer-management systems are connected to one another. Replacing a platform means re-verifying those connections.
- Security and compliance. Companies need audit trails showing who can see what information, who approved what, and what actions were taken.
For this reason, enterprise software is often harder to replace than consumer apps.
AI agents could change that structure.
Work that employees once did by opening a software screen can be carried out by AI that logs into several systems. The value of screens and manual-work features may decline.
But the underlying systems that store data, control access permissions and verify results are still needed.
That AI changes how software is used does not mean all software loses its economic value.
3. Serval's challenge: can it really replace the incumbent?
According to Reuters on Oct. 29, AI agent startup Serval raised $150 million at a $3 billion valuation. That is about triple the $1 billion valuation of December 2025.
Serval's strategy is not to add AI features on top of existing software. It aims to automate companies' IT work while replacing the incumbent platform itself.
Serval CEO Jake Stauch said electric-vehicle maker Lucid moved from ServiceNow to Serval in 45 days.
This is the CEO's claim, and the scope and completeness of the migration have not been independently verified. It should not be taken as evidence that every large company can swap out its systems at a similar speed.
The key question runs deeper than migration speed. Can a new AI company operate data integration, security, access permissions and audit systems at the level of the incumbent?
4. AI could also strengthen existing SaaS
AI agents do not have to be rivals of incumbent SaaS.
Before an AI changes an employee's pay information or alters a critical system setting, a company must check permissions and approval steps. The more work is automated, the more important the platform that controls it may become.
Incumbent SaaS companies already sit inside customers' work data, system connections and security frameworks. Adding AI agents lets customers expand automation without a full platform replacement.
Competition could go in two directions.
| If AI agents weaken SaaS | If AI agents strengthen SaaS |
|---|---|
| Less need for existing work screens | Incumbent platform manages AI work |
| Lower switching costs to new systems | Data and integration become more valuable |
| Price pressure on existing subscriptions | More revenue per customer from AI features |
Which side prevails should show up in customers' actual contracts and usage patterns.
5. What if customers pay for work done by AI, not headcount?
The classic SaaS pricing model is a subscription based on the number of users.
But if AI agents handle many tasks for employees, the number of people who log into the software directly may shrink.
Pricing could change as a result.
- Per user: charged by the number of employees using the software
- Per workload: charged by requests, tasks or transactions the AI handles
- Per outcome: charged by problems solved or tasks completed
This shift is not necessarily bad for incumbents. If revenue from AI automation exceeds the lost user subscriptions, revenue per customer could rise.
Conversely, if customers use AI to do the same work at lower cost, the incumbent's pricing power could weaken.
Outcome-based pricing also brings difficulties: measuring results, assigning responsibility and forecasting customer budgets. The three models may coexist for some time.
Insight Times Editorial Desk
