BIT Research Report | AI Meets SaaS: Who Are the Tears of the Era, and Who Are Turning the Tide Against the Wind?

Null Key Data: Global SaaS Market (2025) approximately $408 billion | SaaS sector market value evaporated by about $2 trillion | IGV Software ETF down approximately 22% year-to-date | Salesforce FY2026 revenue $41.5 billion | ServiceNow Q1 2026 revenue $3.77 billion

1. SaaSpocalypse: Definition of the Historic Event in 2026

In early 2026, Wall Street experienced the largest AI-driven valuation restructuring in software history.

Event Timeline:

January 12: Anthropic releases Claude Cowork, a desktop AI product capable of autonomously executing multi-step workflows across applications

January 30: Anthropic open-sources 11 business plugins covering legal, finance, marketing, sales, and customer support

February 3-5: Market crash. In 48 hours, the SaaS sector lost $285 billion in market value

Core Logic: Traditional SaaS charges per seat. If 10 AI agents can perform the work of 100 employees, a company only needs 10 Salesforce seats instead of 100. Jason Lemkin’s widely circulated quote on Wall Street: “If 10 AI agents can do the work of 100 reps, you need 10 Salesforce seats, not 100.”

Scale of Loss: Total market value loss reaches $1 to $2 trillion (from peak). Thomson Reuters’ largest single-day decline, LegalZoom plunges nearly 20%. The forward P/E ratio of the software sector compresses from about 84x to 22.7x.

2. This is not the end of SaaS, but the beginning of transformation?

Three reasons for optimism:

Proprietary Data Moat: General AI agents cannot replace specialized agents trained on five years of a company’s own CRM data. Salesforce’s data is stored within Salesforce; ServiceNow’s ticket history is within ServiceNow—assets that general AI cannot access.

Underestimated Migration Costs: Replacing deeply embedded enterprise software involves multi-year cycles, millions of dollars in costs, and retraining thousands of employees. SaaStr founder Jason Lemkin points out: Building a functional app with AI programming tools only completes about 2% of the work needed to operate an enterprise software platform.

Compliance and Governance Necessities: In regulated industries like banking, healthcare, and government, enterprise software’s value isn’t just automation—it includes audit trails, compliance records, and access controls. General AI agents currently cannot replace this layer of functionality.

Key Data Refutation: At the peak of selling frenzy, ServiceNow exceeded earnings guidance for the ninth consecutive time, with revenue growth accelerating to 22%. Salesforce recorded $41.5 billion in revenue for the full year. HubSpot maintained 19% growth. These are not figures typical of an industry in collapse.

3. How SaaS Companies Can Counterattack: Three Strategic Pillars

Build proprietary AI agents: Train exclusive agents on their own platform data, rather than waiting for third-party agents to replicate their functions. Agentforce runs on Salesforce CRM data; Now Assist runs on ServiceNow ticket data—advantages that general AI cannot copy.

Transform Pricing Models: Shift from “per seat” to “per result” charging. In Q1 2026, half of ServiceNow’s new business was achieved through non-seat pricing—this is the most critical structural data point in the sector. Goldman Sachs calls this new model “Results-as-a-Service.”

Become an AI Governance Layer: Large enterprises need a trusted platform to unify management, auditing, and security of all AI agents. ServiceNow’s “AI Control Tower” and Salesforce’s “Agentforce Trust Layer” are competing for this key infrastructure role.

4. Key Listed Companies to Watch

5. Salesforce (CRM) — “Agentforce Bet”

FY2026 revenue $41.5 billion, up 10% year-over-year

Agentforce standalone ARR: $800 million, up 169%; over 29,000 contracts signed

RPO (Remaining Performance Obligation): $72.4 billion, up 14%, indicating no customer churn

Approved $50 billion share buyback

Key Point: Will Agentforce be able to drive organic acceleration independently in FY2027 after divesting Informatica’s $1.1 billion contribution?

2. ServiceNow (NOW) — “AI Control Tower”

Q1 2026 revenue $3.77 billion, up 22% (ninth consecutive quarter surpassing guidance)

Now Assist ACV target raised from $1 billion to $1.5 billion, a 50% increase in a single quarter

Renewal rate: 97%, stable for six consecutive quarters

Half of net new business achieved through non-seat pricing

CEO McDermott’s words: “That will exceed $1.5 billion, and we need to raise it by another $500 million or more. It’s incredible.”

Key Point: A benchmark validation of the pricing model transformation, the most valuable data point for sector comparison

3. HubSpot (HUBS) — “Midfield Defender”

2025 full-year revenue $3.13 billion, up 19%; stock price halved by 70-80% from peak

Bull case: Mid-sized enterprise clients are less likely to build AI in-house; HubSpot’s integrated ease of use remains a differentiator

Bear case: Klarna has publicly announced replacing Salesforce contracts with AI; if this trend spreads to mid-sized enterprises, structural pressure becomes unavoidable

4. Workday (WDAY) — “HR Data Moat”

Employee data, payroll, talent records—AI needs Workday’s data for any HR planning

Key Point: Compliance and regulatory requirements make HR software one of the most difficult SaaS categories to replace

5. 2026 Pricing Revolution: The End of the Seat Era

Three current models competing industry-wide:

Consumption-based billing: Charged per query/task/token, more flexible revenue but higher volatility

Results-based billing (Results-as-a-Service): Charged per completed work order, reviewed contract, or generated lead—Goldman Sachs considers this the ultimate form

Hybrid billing: Retain platform access via seat licenses, with additional charges for AI work units—currently the most adopted by enterprises

Most important leading indicator: Who can report a quarter where AI result revenue truly surpasses its replaced seat revenue? This will be a historic data point defining future valuation logic for the sector.

6. Investment Risk Warnings

Not all SaaS will survive: Project management, document tools, simple marketing automation—these are the first repetitive, rule-based tasks to be taken over by AI agents. ERP, HR, compliance infrastructure—due to migration costs and regulatory barriers—are more defensible. Gartner predicts: by 2030, 35% of SaaS single-point tools will be replaced by AI agents, 65% will survive but in altered forms.

Valuation compression may not be over: The software sector’s forward P/E ratio has compressed from 84x to 22.7x, but if disruption outpaces adaptation, further downside remains. Distinguishing “sector cheapening” from “sector should be cheap” is the most critical judgment now.

Build vs. Buy Threats: AI programming tools greatly enhance large enterprises’ ability to develop custom software internally. Klarna’s case is not isolated; it signals a trend worth monitoring.

Summary

Conservative investors can observe sector resilience through ServiceNow and IGV ETF tracks; growth-focused observers should watch whether Salesforce’s Agentforce can sustain organic growth momentum. In 2026, SaaS is no longer about selling seats but about platforms that make enterprises truly indispensable—whether employees are human or intelligent agents.

BIT’s US stock trading connects directly with licensed brokers, covering all core US stocks and ETFs. Supports stablecoin deposits and withdrawals, helping crypto users capture the 2026 AI stock dividends with one click. Services may vary by jurisdiction, some regions (including but not limited to Hong Kong) may be unavailable.

Data as of April 2026. Sources include: Salesforce Inc. (SEC Form 8-K, Feb 25, 2026), ServiceNow Inc. (SEC Form 8-K, Apr 22, 2026), HubSpot Inc. (SEC Form 8-K, Feb 11, 2026), FinancialContent, Taskade, NxCode, Humai Blog, Goldman Sachs “Results-as-a-Service” research report, JPMorgan software sector analysis, Gartner IT spending forecast, Precedence Research, Cirra AI, Fortune, 24/7 Wall St., Redevolution, TechStartups.

Disclaimer: This report is authored by Jun, an invited analyst from BIT’s US stock division, for reference only. The included stocks and ETFs are used as industry case studies and based on publicly available financial reports; it does not constitute investment advice, stock recommendations, or trading inducements. Historical data and institutional forecasts are for reference only and do not predict future market performance or returns. Past performance is not indicative of future results. Investing involves risks, including the potential loss of principal. Clients should consult qualified financial advisors before making any investment decisions.