Salesforce Didn't Invent SaaS. It Hired Actors to Protest a Funeral It Staged for the Old Way.

In March 2000, executives arriving at a Siebel Systems conference walked into a picket line. Protestors with signs and chants. Except the protestors were actors, hired by a 35-person startup operating out of a rented one-bedroom apartment on Telegraph Hill, and the thing they were protesting wasn't a labor dispute - it was the very idea of software you install.⁵ The startup was Salesforce. It would later stage a launch event depicting a literal hell for traditional enterprise software, complete with a fighter jet shooting down a biplane.⁵ None of this was about code. It was theater. And the theater is the whole story.

The official version is that Salesforce invented software-as-a-service. It didn't. Hosted software existed before it, the term came after it, and at least one rival was shipping web-delivered subscription CRM at the same time. What Salesforce actually invented was the argument.

Here is the thesis a smart friend can repeat at dinner: Salesforce didn't pioneer the technology of SaaS so much as it pioneered the belief in it. The real moat was evangelism. The company took a delivery method that buyers found abstract and confusing and turned it into a movement with a villain, a slogan, and a fighter jet.

What Salesforce actually built - and what it didn't

Strip away the marketing and there is a genuine technical bet underneath, and it matters. The hosted-software companies of the late 1990s - Application Service Providers - ran a copy of the software on dedicated infrastructure for each customer. One tenant, one stack. That meant every customer was a small, expensive island to maintain. Salesforce instead built a multi-tenant architecture: one set of shared code and infrastructure serving every customer at once, with their data logically separated.⁸ Crucially, the company never transitioned from a prior model. There was no CD-ROM era to abandon, no on-premise legacy to migrate off. It was built as pure web-delivered subscription software from day one.⁸

That single architectural choice is what made the economics work. When the next customer costs almost nothing to onboard because they share the same machines and the same upgrade, your gross margin starts to look like software and your operating leverage compounds. But here is the correction that the founding legend skips: Salesforce was not the first to deliver software over the internet. NetSuite was founded in 1998, and other hosted-CRM rivals were live around the same time. The defensible claim is narrower and truer - Salesforce was the first pure-play, multi-tenant SaaS company built from scratch at scale.⁸

Why the marketing was the product

Multi-tenant architecture is invisible. No buyer ever fell in love with logical data separation. The problem Salesforce faced in 1999 was not that its technology was weak - it was that enterprise buyers had no mental category for what it was selling. 'Software as a service' was not yet a phrase a CIO had heard; it consolidated in analyst white papers and industry usage in the years that followed, not at Salesforce's founding. So Salesforce did something more clever than coining a term. It picked a fight.

The 'No Software' campaign gave the abstract a shape. A red circle with 'software' crossed out. A theatrical hell for the installed-software way of life. A fighter jet downing a biplane to dramatize new versus old.⁵ And the actors picketing outside Siebel's conference - reported in The Wall Street Journal in March 2000 - did the most important work of all.⁵ They named a villain. Once buyers had a villain (slow, expensive, install-it-yourself enterprise software) and a banner (No Software), they suddenly had a category to belong to. The geography of the argument did the selling. Benioff understood enterprise software from the inside - he had spent 13 years at Oracle, where he was the youngest vice president in company history - and he knew exactly which incumbent fears to weaponize.⁶

There is a delicious irony here that proves the point. Salesforce is software. It is nothing but software. The slogan was technically false and strategically perfect, because it wasn't describing a product - it was describing a side. You weren't buying a CRM database. You were defecting from the old world.

From a Telegraph Hill apartment to a public referendum

The bet took five years to face a real verdict. Salesforce was incorporated on March 8, 1999, by four co-founders - Benioff, Parker Harris, Frank Dominguez, and Dave Moellenhoff - working from a rented one-bedroom unit at 1449 Montgomery Street.¹ By late 2003 it was ready to go public, and the road was bumpy: it filed its S-1 in December 2003, the SEC balked over how it accounted for sales commissions, and Benioff broke the quiet period, pushing the debut back by roughly a month.³

On June 23, 2004, Salesforce listed on the NYSE under the ticker CRM, offering 10,000,000 shares at $11.00 to raise $110 million.² Note what that price already tells you: the offering had been ranged at $7.50 to $8.50 and was pushed above range before pricing - the popular '$8 IPO' shorthand describes a range that never happened.⁴ The stock closed its first day at $17.20, a gain of more than 55%.⁴ The public had bought the argument.

But wasn't the real win the technology, not the theater?

The honest objection is that this overrates the marketing. Picket lines and slogans get you attention; they don't get you to $1 billion in revenue in 2009 - which made Salesforce one of the first cloud companies to reach that milestone - or to a record $41.5 billion in its fiscal 2026.⁷ Surely the multi-tenant architecture, not the fighter jet, built that. And the architecture was genuinely load-bearing: it is what let each new customer cost almost nothing to serve, the engine of the whole economic story.⁸

True - but it confuses the necessary with the decisive. The architecture was necessary; plenty of competitors had hosted software, and at least one rival had multi-tenant ambitions of its own. What Salesforce had that they didn't was distribution for the idea. The theater is what got enterprise buyers to take a phone call about a category they'd never heard of, from a startup in an apartment, against incumbents they'd trusted for a decade. The architecture made the model profitable; the evangelism made the model believed. Take away the architecture and the economics collapse. Take away the marketing and a superior product dies in obscurity while a slower-moving incumbent keeps the account. Both were required. Only one was rare.

That is the lesson worth keeping. Salesforce's most-copied invention was never multi-tenancy - that became table stakes. It was the discovery that when you are selling something the market cannot yet picture, the product is the argument, and the moat is how many people you can get to repeat it. Salesforce hired actors to protest the old way, and then it spent twenty-five years making the protest come true. The slogan was a lie about the product and the truth about the company. It really did sell No Software. It sold a side.