The fear that AI agents will take application software's revenue did not show up in the third-quarter results of the nine application software companies SaasBuzz covered this season. Their median revenue growth was 8.7% from a year earlier, against 8.5% in the second quarter, and six of the nine raised or held their full-year revenue forecasts, by a SaasBuzz analysis of their results releases. The three that trimmed cut their midpoints by less than 1%, and none of them blamed AI.
The forward-looking numbers are softer. Four of the five companies that report a contracted backlog or annual recurring revenue (ARR) in their releases saw that measure grow more slowly than revenue, by half a point to a point; at Tyler Technologies, ARR grew in line with recurring revenue. ServiceNow (NYSE: NOW), the largest of the nine, forecast that its 12-month backlog would grow 18.5% in the fourth quarter, after 20.5% in the third. That is a cooling worth watching, but it is a small one, and so far it is not the collapse the bear case describes.
The test matters because the shares moved first. ServiceNow's stock fell 27% on October 29, the day after its results, Seeking Alpha reported, and our count of October's software stocks found the month turned in its final week. Part of the argument against the sector is that agents doing the work will need fewer software seats.
What the nine reported
Each company's revenue growth in the quarter to September, the quarter before, and its full-year forecast against the one it gave three months earlier:
Company | Q3 | Q2 | Full-year guide |
|---|---|---|---|
ServiceNow | 20.5% | 24.0% | Raised |
Twilio | 19.9% | 22.0% | Raised |
Commvault | 13.1% | 11.4% | Raised |
Asure | 10.4% | 23.2% | Raised |
Tyler | 8.7% | 8.2% | Trimmed |
Rimini Street | 8.3% | 6.7% | Held |
Coveo | 8.3% | 8.5% | Trimmed |
GoDaddy | 6.0% | 6.6% | Held |
SPS Commerce | 4.1% | 5.6% | Trimmed |
Median | 8.7% | 8.5% | 6 of 9 |
Source: SaasBuzz analysis of the companies' results releases. Total GAAP revenue against the year-earlier quarter, calculated from the reported figures; Commvault's and Coveo's fiscal quarters to September 30 and June 30 are shown as Q3 and Q2. The guide compares the full-year revenue midpoint (ServiceNow and Commvault: subscription revenue; Twilio: reported growth) with the previous one. Rimini Street reiterated a "Rule of 20" target, revenue growth plus adjusted EBITDA margin, which in July it paired with 4% to 6% revenue growth.
Five of the nine grew more slowly than in the second quarter, two grew faster and two were within half a point. The largest drop, Asure Software's, came after a 23% second quarter; Asure said third-quarter revenue rose 9% organically and that its raised forecast "incorporates the recent acquisition completed in October 2026." The raises were small too: Commvault lifted its subscription revenue range to $1,127 million to $1,133 million from $1,119 million to $1,129 million, and Twilio raised its reported growth outlook to 20%, from 18% to 18.5%.
The companies that sell the AI build-out are a separate story. Microsoft's revenue grew 19%, Alphabet's 25% and Amazon's 12% in the same quarter, by their releases, and our analysis of the October 28 results covered how Microsoft's and Alphabet's cloud businesses compare with subscription software. They are left out of the count here, as are Fortinet and F5, security and networking companies with large product revenue lines.
Where growth is slowing
Revenue is a lagging number: most of a subscription company's quarter was signed months or years earlier. If agents were starting to replace seats, the first sign would be in what companies have under contract for the year ahead. That is where the softness is.
Company | Measure | Growth | Revenue |
|---|---|---|---|
ServiceNow | cRPO | 20.5% | 21% |
Commvault | Sub. ARR | 20% | 21% |
Tyler | ARR | 11.0% | 11.0% |
Rimini Street | Adj. ARR | 9.8% | 10.7% |
GoDaddy | ARR | 5.1% | 6.0% |
Source: the companies' third-quarter results releases, growth as each reports it, against the closest revenue line: subscription revenue for ServiceNow and Commvault, recurring revenue for Tyler, revenue excluding PeopleSoft support for Rimini Street, total revenue for GoDaddy.
The gaps are small and the measures differ, so the five are not one trend. The clearer signals are in two forecasts. ServiceNow expects current remaining performance obligations (cRPO), the contracted revenue it will recognize over the next 12 months, to grow 18.5% in the fourth quarter. Twilio (NYSE: TWLO) expects organic revenue growth, which leaves out acquisitions and increases in carrier fees it passes on at cost, of 11% to 12%, after 16% in the third quarter.
Both companies have guided low before. Twilio's forecast for the third quarter was also 11% to 12% organic growth, its August release shows, and it delivered 16%. ServiceNow had guided third-quarter cRPO growth to 19.5% and reported 20.5%. On the October 28 call, Gina Mastantuono, ServiceNow's president and chief financial officer, attributed part of the fourth-quarter slowdown to timing. Early renewals "contributed about half of the beat in Q3," she said, and "we pulled those deals out of Q4 and into Q3," according to a transcript published by The Motley Fool. Our report on the quarter sets out the other factors she named.
None of the three trims was put down to AI. Tyler Technologies said it had adjusted its revenue guidance "to reflect shifts in the timing of new sales during the year," a 0.2% change at the midpoint, its prepared remarks say. SPS Commerce said its third-quarter growth was held back by the sale of a revenue recovery business, which its July forecast already allowed for, and "variability in the 1P Revenue Recovery take rate model." Coveo narrowed its total revenue range to $160 million to $162 million while raising the bottom of its SaaS subscription range.
The case that AI is eating software
The bear case does not need revenue to fall this year: as agents take over tasks, customers buy fewer seats at renewal, and the damage arrives on a delay. Some of its best support comes from the other side.
Gartner, in an article published October 30, said 91% of IT leaders anticipate disruption to current SaaS models, and that 51% are reassessing "point solutions," in its words, against 22% reconsidering major systems such as ERP and CRM. Its advice to buyers includes: "Negotiate reduced seat-count commitments." Bank of America analyst Tal Liani, quoted by Seeking Alpha, said the focal points of ServiceNow's results were "the 120bp and 40bp shortfall in 4Q in cRPO and subscription growth guidance," and added: "We don't fully subscribe to management's explanation on the quarter's weakness." He remained constructive on the stock.
ServiceNow itself is hedging its pricing. On the call, Bill McDermott, its chairman and chief executive, said "the customer likes the predictability of the seat-based pricing model for their budgeting purposes," and added: "And also, we're driving consumption."
What Gartner and ServiceNow's CFO argue
Gartner's case is that the systems agents need are the ones they are supposed to replace. The firm said "there are currently no viable AI-native providers capable of replacing major enterprise systems like an ERP at large, global organizations at scale."
The paradox is that today's AI agents depend on the very systems of record organizations are trying to escape.Patrick Connaughton, vice president analyst at Gartner, October 30
Mastantuono made the company's version of that argument in an interview with Seeking Alpha, published October 31 under the headline "ServiceNow CFO says AI disruption to enterprise software narrative is false." "Enterprises are deploying at scale, and our momentum continues to accelerate and be strong, but at the same time demand for our core workloads remains strong," she said. "AI is pulling our core in a strong way."
Both are views from parties with an interest: Gartner's article is advice to the buyers of this software, and ServiceNow sells it. The results above are the check on both, and for one quarter they sit closer to Gartner's version than to the selloff's.
What it means for buyers and investors
For buyers, the Gartner advice and McDermott's answer point the same way: seat counts are becoming negotiable, and vendors are adding consumption pricing beside them.
For anyone weighing the companies, this quarter's numbers rule out the fast version of the story, in which AI cuts into revenue already under contract. They cannot rule out the slow version, because the slow version shows up first in backlog and ARR growth, and those measures are where four of the five companies that report one came in below their revenue growth.
What would change the conclusion
ServiceNow's fourth-quarter cRPO against its 18.5% forecast
Twilio's fourth-quarter organic growth against 11% to 12%, after it beat the same range in the third quarter
Whether backlog and ARR growth fall further behind revenue growth, or catch up
Results from companies that price mainly by the seat, which report later in the season and are not in this sample
Whether the 22% of organizations Gartner found reconsidering major systems moves toward the 51% reassessing niche tools
Method
The analysis covers every application software company in SaasBuzz's earnings ledger that reported between October 27 and October 30, 2026: ServiceNow, Tyler Technologies, Twilio, GoDaddy, Coveo, Rimini Street, SPS Commerce, Asure Software and Commvault. The sample is the companies SaasBuzz covered, not all of software, and nine companies cannot carry a claim about the whole sector. Microsoft, Alphabet and Amazon (infrastructure), Fortinet and F5 (security and networking) are shown apart; Reddit and DoubleVerify (advertising) and Oracle, which reported no quarter, are left out.
Revenue growth is total GAAP revenue against the year-earlier quarter, calculated from each release's income statement, and includes acquisitions. Forward measures are each company's own, as reported, and are not comparable across companies. Guidance compares midpoints of the latest full-year range with the previous one. The calculation is saved with this article's working files.
