On February 22nd, 2000, twenty-five actors in red T-shirts marched outside the Siebel CRM user conference in San Francisco carrying signs that read "End of Software". A fake news crew from a fake station called Channel 22 interviewed them like it was a political movement. Siebel, the on-premise CRM giant of the day, called the police , which was the cue every business reporter in the room needed to file the story. The startup that paid for the actors was a year-old company called Salesforce, and within ten years it had eaten the category whole.
That was the first rebrand.
The next seven followed the same shape. End of Software became Cloud, Cloud became Social (Chatter), Social became Mobile (Salesforce1), Mobile became Customer 360, Customer 360 became Einstein, Einstein became Agentforce. Each rebrand attached the next industry-defining noun to the platform — and, more importantly, expanded what a Salesforce seat was supposed to do. You weren't paying for CRM, you were paying for the system of record. You weren't paying for the system of record, you were paying for the customer graph. You weren't paying for the customer graph, you were paying for the AI that ran on it. The price went up. The seat count went up. The narrative absorbed every shift the rest of the industry tried to use against it.
In March 2026, at TDX in San Francisco, Salesforce announced its eighth rebrand. It is called Headless 360 . It is the first one that does not extend the moat.
Headless 360 makes every Salesforce capability — CRM, service, marketing, commerce, Slack — accessible directly as APIs, MCP servers, and CLI commands. The browser is optional. Agents in Claude Code, Codex, Windsurf, or VS Code can read and write Salesforce data without a human ever logging in. The pitch, from EVP Madhav Thattai, is that Salesforce wants to "expand the tent" and bring in developers who never touched the platform before.
The inconvenient corollary, said quietly, is that if the browser is optional then so is the seat.
(For readers who don't work in SaaS: Salesforce's revenue model is built almost entirely on per-user, per-month seat licences. The price is anchored to a human logging in and clicking around in a Salesforce-built UI. When the UI moves out of the browser and into an agent that one human is operating on behalf of fifty workflows, the relationship between value delivered and seats consumed breaks. The vendor that figures out how to charge for that gets paid. The vendor that doesn't watches its growth flatline while the work the platform is doing keeps going up.)
The numbers under the announcement say the market has already noticed. Salesforce stock is down roughly 20–30% year-to-date in 2026 while the S&P 500 is up about 4%. Subscription growth is decelerating: 10% in FY25, guided to 8.5% for FY26. Agentforce, the product Marc Benioff describes as the fastest-growing in Salesforce's 26-year history , reached $800m in ARR by the end of FY26 — up 169% year-on-year, a genuinely impressive ramp, and still 1.9% of the company's $42.7bn total ARR.
The market priced the transition before Salesforce narrated it.
The reason it is happening now, and not five years ago, is that the architectural argument has finally been made out loud. Satya Nadella, on the BG2 podcast in late 2024 , described traditional business applications — CRM, ERP, ticketing — as CRUD databases with business logic stapled on top, and predicted that the logic layer would migrate up into the AI tier. The data layer underneath becomes infrastructure: still important, increasingly fungible. Jason Lemkin at SaaStr has been more pointed about the operational reality. His own organisation, he writes, spends roughly fifteen times more on AI agents than on Salesforce licences , and his read on Headless 360 is that it isn't a product announcement — it's "a description of what's already happening." The Klarna saga — the viral 2024 claim that Klarna had ripped out Salesforce in favour of an in-house LLM, walked back by the CEO six months later — is the same signal in a different costume. The rumour broke before the revenue did. That the story was plausible enough to spread is the data point.
It's worth naming where the analogy stops being useful before going further. Salesforce is not dying next quarter. The company has $42.7bn in ARR, somewhere around 150,000 customers, two decades of integrations the rest of the enterprise stack is wired into, and the data gravity that buys every successful platform vendor a final decade of life. Agentforce is real revenue, growing fast. Benioff has earned the right to mock the SaaSpocalypse framing on the earnings call, and on the call he is correct. None of the things above suggest collapse. They suggest something more specific.
I have been close enough to this platform for long enough to feel the difference. I have shipped on Salesforce across most of the rebrands listed above, written the integrations the marketing material promised would be unnecessary, and watched orgs spend years untangling the architectural debt each rebrand introduced before the next one landed. The pattern is real, and so is the muscle memory. What is different this time is that the rebrand is asking customers to do less inside Salesforce, not more — to think of it as a backend, not a product. That is not the move you make from strength.
The honest question is not whether Salesforce survives. It is what the ratio of seat-based to consumption-based revenue at the company looks like in five years, and whether Salesforce can complete that transition before the layer above it — the agent layer Headless 360 is now politely escorting in — turns the layer underneath into a commodity. The answer is genuinely uncertain. Salesforce has the data, the trust footprint, and a CEO with more rebrands left in him than most of the industry combined. None of that is nothing.
But the writing on the wall is not "Salesforce ends." It is "the Salesforce that rebrands its way out of every cycle ends here." What replaces it on the same revenue base is the open question, and for the first time in 25 years, the answer is being written by people who don't work there.
No comments yet