Large-cap application SaaS in the AI era Report Interpretation

Overview

This North American software earnings review assesses whether generative AI threatens or reinforces large-cap application SaaS. Morgan Stanley concludes that recent results offer early support for the AI bull case, but that sustained gains require visible AI revenue, resilient margins, and a clearer path to organic growth reacceleration.

Core views

Morgan Stanley argues that CY2Q26 earnings softened the “SaaS is dead” thesis even though estimate revisions remained limited. Large-cap application SaaS had underperformed the S&P 500 by roughly 20 percentage points over the prior year, and companies with FY1 revenue revisions of only plus or minus 1% outperformed the broad software group by an average 18% in the five days after CY2Q26 results. This contrasts with a 4% underperformance after comparable CY2Q25 revisions a year earlier. The institution attributes part of the rebound to depressed starting valuations—average EV/FCF and P/GAAP earnings multiples were about 60% below their three-year averages—but views the more important change as reduced confidence in the AI bear case. It cautions that earnings did not yet establish a fundamental reacceleration case. The first pillar is application control. Morgan Stanley argues that models supply probabilistic reasoning, while enterprise software retains governed data, permissions, business logic, and reliable execution workflows. Agents may change the user interface, but applications can remain the deterministic execution layer and therefore avoid economic disintermediation. Supporting evidence includes Salesforce’s disclosure that nine of its top 10 AI companies use Salesforce and Slack, with cohort spend up 435% year on year in 2Q; ServiceNow’s AI Control Tower reaching more than 50 live customers within six months; Atlassian’s MCP calls rising 400% in the quarter and combined MCP/CLI users exceeding 1 million; Workday having more than 5,500 customers using at least one organic agent; and Intuit reporting millions of users of AI-native experiences. The second pillar is less seat dependence. The report argues that falling logins or seats would not necessarily mean a lower addressable market, because agents can generate API calls, model-context-protocol calls, data requests, actions, and completed workflows without human users opening an application. Vendors may monetize platform access, consumption, data, agent actions, and outcomes rather than seats alone. Importantly, broad seat contraction has not appeared in 2Q results. Salesforce reported year-on-year growth in Sales, Service, and Slack seats alongside 3.2 billion Agentic Work Units, up 97% quarter on quarter; ServiceNow said 50% of net new business is already non-seat based; and Workday signed 200 customers to Flex Credits in Q2. Morgan Stanley nevertheless identifies a transition risk: moving from seat-based pricing to platform-plus-consumption pricing can initially pressure ACV before usage ramps. Third, the report sees inference leverage as a potential source of application-layer differentiation and margin support. As foundation and open-weight models become more interchangeable, software platforms with proprietary data, workflow context, permissions, and execution environments may capture more value by choosing models, routing tasks, and delivering outcomes. Lower inference costs may also reduce AI cost of goods sold, though the key question is how much savings vendors retain. ServiceNow cites more than 20 years of operational intelligence across 7 trillion workflows; Atlassian says its Teamwork Graph contains 25 years of work data and more than 200 billion objects and connections, enabling up to 44% more accurate answers with 48% fewer tokens; Workday emphasizes its unified data and security model; and Intuit points to proprietary financial data and tax/accounting-specific models while committing to company-level margin expansion. Fourth, Morgan Stanley argues AI can expand software’s addressable budget from IT spending toward labor spend, provided enterprises can redesign workflows, establish governance, and demonstrate return on investment. Near-term monetization is likely to come through premium tiers and bundles before consumption and outcome pricing become more meaningful. Salesforce’s agentic SKU carries a 60–80% premium to traditional seats but is only about 5% penetrated. ServiceNow’s AI ACV exceeded $1 billion, with deals containing five or more AI products up 5.5x year on year and a $1.5 billion AI ACV target. Workday’s AI SKUs reached nearly $600 million ARR, up more than 200% year on year and more than 20% quarter on quarter. Intuit’s outcome-oriented AI experiences are positioned against a nearly $200 billion TAM that is only 7% penetrated. Fifth, embedded distribution is presented as an incumbent advantage. Existing customer relationships, contracts, integrated data, security permissions, and tailored workflows can reduce deployment friction and speed agent adoption. Salesforce’s Agentforce ARR, including Slackbot/Headless 360, grew 240% year on year to more than $1.5 billion, and the report cites a six-week Uber for Business launch. ServiceNow has 658 customers generating more than $5 million in ACV; Atlassian is present in 85% of the Fortune 500 and serves more than 350,000 customers; Workday’s Deployment Agent is live with more than 4,600 customers and nearly 24,000 users; and Intuit has more than 150,000 accountants on Intuit Accountant Suite. Within applications, Morgan Stanley continues to prefer infrastructure and security software overall because applications are later-cycle beneficiaries. It remains constructive on ServiceNow after a cleaner and larger Q2 beat in cRPO and subscription revenue, easier second-half comparisons, potential federal-demand support, and a possible 2027 AI-capacity growth driver. It reiterates a positive view on Atlassian after accelerating cloud growth, a strong FY27 cloud guide, rapid growth in agentic usage, and medium-term monetization potential from capacity overages. Salesforce is incrementally better following modest Q2 cRPO beats and Agentforce ARR growth, but 7% organic constant-currency subscription-revenue growth, down from 8%, means the report still seeks evidence of durable acceleration. Intuit faces uncertainty after a strategic shift in its TurboTax DIY business amid competition. Workday’s AI traction counters the bear case and FY28 subscription-revenue guidance indicates possible stabilization, but monetization in critical HR and finance workflows is expected to take time, leading the report to fade near-term outperformance.

Analysis framework

Morgan Stanley compares post-earnings estimate revisions, share-price reactions, and valuation levels, then tests five proposed AI-era durability pillars against company-specific 2Q operating evidence. It assesses whether AI usage is supporting application control, consumption monetization, inference economics, budget expansion, and distribution advantages, and then translates those findings into differentiated views on the covered companies.

Methodology notes

• Competition & strategyEconomic Moat and Competitive Advantage

  Moat & Journey framework

  The report evaluates whether proprietary enterprise data, workflow control, customer relationships, and distribution create durable competitive advantages, and whether each company has a credible path to realizing AI-driven growth.

• Industry AnalysisVolume-price decomposition

  Seat-based versus consumption, premium, and outcome-based monetization

  The analysis separates traditional seat growth from pricing, consumption, agent actions, and outcome-based revenue to assess whether AI expands rather than cannibalizes software demand.

• Valuation methodsEV/EBITDA valuation

  EV/FCF, P/GAAP earnings, P/E, PEG, and discounted FCF valuation references

  The report uses relative valuation multiples and company-specific discounted cash-flow or free-cash-flow assumptions to frame valuation, while noting that large-cap SaaS entered earnings at materially depressed multiples.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• ServiceNow (NOW)

  Preferred application-platform exposure supported by a durable moat, improving growth setup, and expanding AI-capacity monetization.

  Strengths

  Cleaner and larger Q2 cRPO and subscription-revenue beat; easier 2H26 comparisons; more than $1B AI ACV; early federal-demand support; differentiated operational context.

  Comparison

  Morgan Stanley remains constructive relative to the less-convicted near-term views on Workday and Intuit.

  Risks

  Competition from low-end vendors and SaaS peers; sales-productivity pressure if new markets prove challenging.

• Atlassian (TEAM)

  Preferred application-platform exposure with accelerating cloud growth and agentic-workflow opportunity.

  Strengths

  Strong FY27 cloud guide, Teamwork Graph context advantage, rapid AI usage growth, and medium-term opportunity to monetize capacity overages.

  Weaknesses

  Management has not yet monetized excess AI-related usage.

  Comparison

  The report reiterates a positive view as seat concerns fade and AI progress becomes clearer.

  Risks

  AI-driven productivity could pressure enterprise headcount; AI-native and in-house alternatives; pricing-model transition.

• Salesforce (CRM)

  Incrementally more positive following an improving rate of change in upside to guidance.

  Strengths

  Modest Q2 cRPO beats, Agentforce ARR of more than $1.5B, and a 60–80% agentic-SKU price premium.

  Weaknesses

  Organic constant-currency subscription-revenue growth moderated to 7% from 8%, and durable acceleration remains unproven.

  Comparison

  Screens incrementally better on Morgan Stanley’s Moat & Journey framework but has a longer reacceleration journey than preferred platform peers.

  Risks

  Agentforce and Headless initiatives may fail to stabilize growth; AI investment may weigh on margins; alternative platforms may crowd out front-office demand.

• Intuit (INTU)

  AI-enabled platform opportunity, but Morgan Stanley is less convinced near term after a strategic shift in TurboTax DIY.

  Strengths

  AI-native experiences, cross-platform monetization, a growing mid-market business, and embedded accountant distribution.

  Weaknesses

  New-customer strategy remains unproven and is not expected to be resolved until F3Q earnings in May 2027.

  Comparison

  Less favored near term than ServiceNow and Atlassian.

  Risks

  TurboTax competitive pressure and share loss; slower QBO Advanced adoption; macro sensitivity in Credit Karma and Mailchimp.

• Workday (WDAY)

  AI traction challenges the bear case, but revenue monetization is expected to lag adoption.

  Strengths

  More than 5,500 customers using organic agents; AI products drove more than $100M of new ACV; FY28 subscription-revenue guide of 11% may signal stabilization.

  Weaknesses

  AI adoption in critical HR and finance workflows takes time, delaying growth contribution.

  Comparison

  Morgan Stanley is less convicted near term and would fade near-term outperformance.

  Risks

  Lack of operating leverage and continued growth deceleration.

Key data

• Post-earnings share-price reaction18% average outperformanceLarge-cap application SaaS companies with CY27 revenue revisions of +/-1% outperformed the broad software group in the five days after CY2Q26 results.
• Comparable prior-year reaction4% underperformanceComparable CY26 revisions after CY2Q25 results a year earlier.
• Large-cap SaaS valuation discount~60% below three-year averageAverage EV/FCF and P/GAAP earnings multiples entering earnings season.
• Salesforce Agentforce ARR$1.5B+; +240% YoYIncludes Slackbot/Headless 360.
• ServiceNow AI ACV>$1BThe report says the company is on a path toward its $1.5B AI ACV target.
• Workday AI SKU ARRNearly $600M; +200%+ YoY, +20%+ QoQEvidence of AI product traction, although broader monetization is expected to take time.
• Atlassian Teamwork Graph25 years of work data; >200B objects and connectionsThe report says grounded agents can produce up to 44% more accurate answers while using 48% fewer tokens.

Impact & implications

The report’s central implication is that AI need not displace established application SaaS vendors if they retain control of enterprise context and execution. However, valuation recovery driven by sentiment and depressed multiples must be followed by observable AI revenue, stable or improving margins, and organic growth acceleration to become durable.

Risks

• A shift from seat-based to hybrid platform-and-consumption pricing can create near-term ACV headwinds before consumption usage ramps.
• AI product adoption or monetization may fail to produce sufficient organic growth reacceleration.
• AI investments could pressure margins if inference-cost savings are not retained.
• Competitive AI-native alternatives, alternative platforms, and in-house development could weaken incumbent demand.
• Enterprise deployment of AI in critical workflows may take longer than expected.

What to watch

• Whether AI revenue and consumption growth rise alongside stable or improving gross margins.
• Evidence that AI usage converts from product adoption into capacity purchases, premium tiers, and outcome-based monetization.
• ServiceNow’s cRPO trajectory, federal-demand backdrop, and AI-capacity bundle demand.
• Atlassian’s cloud growth, agentic capacity usage, and eventual monetization of overages.
• Salesforce’s organic subscription-revenue growth and whether Agentforce helps offset legacy-portfolio drags.
• Intuit’s progress in attracting new TurboTax DIY customers by F3Q earnings in May 2027.
• Workday’s pace of AI adoption and the timing of a meaningful revenue contribution.