Salesforce, ServiceNow, Atlassian, and Monday.com: How CRM, NOW, TEAM, and MNDY Survive the Agentic AI Era
Four enterprise software franchises that defined the last software decade entered the summer of 2026 carrying the same wound. Salesforce trades near $152, down roughly a third on the year and off 38% from where it stood twelve months ago. ServiceNow, having executed a five-for-one split in December, changes hands around $95 after surrendering more than half its value from the 2025 peak. Atlassian sits in the low $80s, down some 45% year to date and roughly 74% off its one-year high. Monday.com, the smallest of the group, has lost about 73% over a year and trades in the mid-$60s against a 52-week high near $317.
None of these is a broken business. All four grew revenue at double-digit rates through the drawdown. The market was not punishing failure. It was pricing a single question, and pricing it brutally: what is a software seat worth when the worker sitting in it is an agent?
The Seat Is the Whole Argument
The modern SaaS fortune was built on a simple unit. One license, one human, billed annually, expanded as headcount grew. That model compounded for fifteen years because enterprises kept hiring knowledge workers and kept buying them tools. Agentic AI inverts the assumption. If an autonomous agent resolves the service ticket, qualifies the lead, or closes the sprint, the seat it replaces does not renew. The bear case against this entire cohort is therefore mechanical rather than cyclical. It does not require a recession or a competitor. It requires only that AI works.
Wall Street has been explicit. Bank of America placed a $160 target and an Underperform rating on Salesforce, framing the risk not as a soft quarter but as structural erosion of seat-based growth. That is the sentence haunting every name on this list. The companies that re-rate will be the ones that prove their pricing no longer depends on counting humans.
Salesforce (CRM): The Pivot Is the Position
Salesforce is the most exposed name precisely because it is the most synonymous with the seat. It built a business approaching $42 billion in annual revenue selling licenses one human at a time, and the market now doubts the foundation. Fiscal 2026 closed with revenue up about 10%, remaining performance obligations above $72 billion, and operating cash flow north of $15 billion. The growth is real but decelerating, and deceleration is exactly what the bears expect from a seat model under pressure.
Benioff’s answer is Agentforce, and the conviction behind it is not subtle. Agentforce and Data 360 pushed combined AI and data annual recurring revenue past $2.9 billion, growing more than 200%, with the platform processing close to twenty trillion tokens. The company committed roughly $300 million to Anthropic token usage in 2026 and is rebuilding its price book around completed work units rather than occupied seats. Management still points to $63 billion in revenue by fiscal 2030. The thesis is clean: charge for outcomes the agent delivers instead of chairs it empties. Salesforce is the slowest ship to turn and the most identified with the old model, but it is also furthest along in building the replacement. The stock is a bet that the pivot lands before the seat erodes.
ServiceNow (NOW): The Workflow Layer Was Never a Seat
ServiceNow is the relative winner of this group, and the reason is architectural. It does not sell a chair at a desk. It sells the workflow layer that routes work across an enterprise, and that layer becomes more valuable, not less, when agents start executing the tasks moving through it. First-quarter subscription revenue rose 22% to roughly $3.7 billion, with current remaining obligations up a comparable amount and full-year subscription guidance lifted toward $15.7 billion.
The defining number is pricing mix. Roughly half of new business is no longer seat-based, and customers spending over a million dollars annually on Now Assist grew more than 130%. McDermott’s framing of the platform as the control tower and semantic layer for enterprise AI is marketing, but the underlying claim survives scrutiny: orchestration is the one thing agents cannot disintermediate, because someone has to govern them. The catch is valuation. A price-to-earnings multiple near fifty leaves no margin for a stumble, and any deceleration in Now Assist adoption will be punished without mercy. ServiceNow has the best answer to the seat question and the least room for error.
Atlassian (TEAM): The Market Priced a Disruption the Numbers Refuse to Show
Atlassian is the contradiction in the group. It has fallen the hardest among the large caps, yet its own results argue against the thesis driving the selloff. Third-quarter revenue grew 32% to $1.8 billion, cloud crossed $1.1 billion, and non-GAAP earnings beat estimates by roughly a third. The Service Collection passed $1 billion in annual recurring revenue. Most directly, management stated there is no sign of seat compression, and that customers using the Rovo AI platform expand their spending at roughly twice the rate of those who do not.
That is the opposite of disruption. Rovo credit usage is climbing more than 20% month over month, and the Teamwork Graph gives Atlassian a structural data advantage in feeding context to agents. The stock trades around 3.8 times sales against a sector closer to seven, a multiple that prices terminal decline for a company still accelerating. Founder control under Cannon-Brookes adds a long-horizon discipline the market is currently ignoring. If seat compression never materializes, Atlassian is the clearest mispricing on the list. The risk is that the market is early rather than wrong.
Monday.com (MNDY): Small, Cheap, Cashed Up, and Binary
Monday.com is the highest-beta expression of the entire debate. As the smallest and least entrenched name, it is the most exposed to seat erosion and the most leveraged to getting the transition right. First-quarter revenue grew about 24% to $351 million, AI products already contributed roughly 10% of net new annual recurring revenue, and the company is rebuilding around agent-based automation, hybrid pricing, and a marketplace for AI agents, reinforced by the acquisition of One AI. Margins compressed as the company spent into the shift, with net margin slipping toward 8%.
The balance sheet is the floor under the story. Monday holds roughly $1.2 billion in net cash, near 30% of a market capitalization around $4 billion, and repurchased more than $550 million of stock in a single quarter. Down 73% in a year, it trades at a fraction of past multiples while still growing and self-funding. This is the option in the group: cheap, well capitalized, and binary. If the AI-first rebuild converts its base, the re-rating is violent. If agents commoditize lightweight work management, no balance sheet saves a shrinking unit.
The Verdict
The market has compressed four distinct businesses into one trade, and the trade is wrong to treat them identically. ServiceNow sits closest to safety because it sells orchestration, not occupancy, though its multiple already assumes that truth. Atlassian is the sharpest value if its own data is believed over the tape. Salesforce is the pivot bet, the largest and slowest, wagering that work-unit pricing arrives before the seat dissolves beneath it. Monday is the lottery ticket with a balance sheet, cheap enough that survival alone would pay.
The seat is not being eliminated. It is being repriced, and the repricing will sort the winners from the legacy in the space of about eight quarters. The companies that learn to bill for what the agent finishes will keep their franchises. The ones still counting humans will watch the market count them out.