Anthropic's IPO filing shows sales and marketing spend up 26x in six months

Anthropic's S-1 filing for its upcoming IPO shows sales and marketing spend climbing from $120 million in the first half of 2025 to $3.17 billion in the first half of 2026, a roughly 26-fold jump that outpaced even the company's fast-growing revenue. The filing, first dissected by SaaStr and separately by Digiday, puts a specific number on something software marketers have mostly sensed from the outside: the AI labs are now spending like consumer tech giants to win a market that barely existed three years ago.

What the filing shows

Anthropic's narrower advertising line rose from $33 million in 2024 to $147 million in 2025, which on its own already exceeds Duolingo's entire 2025 sales and marketing budget of $125.7 million, according to Digiday's reporting. The broader sales, marketing and partnerships category, which also covers go-to-market staff and what Anthropic calls Applied AI teams, went from about 10% of revenue to about 19% in the space of a year. Revenue grew fast too, from $1.2 billion to $16.3 billion across the same two half-year periods, but the spending line still grew faster. As recently as the 2023 fiscal year, the filing described advertising spend as too small to disclose.

Anthropic attributes the jump to heavier media and public relations spending behind a string of launches: Claude Code's debut in February 2025, a run of model releases through Opus 4.5 in November 2025, and the January 2026 launch of Claude Cowork. A large share of the broader total also flows through channel partners rather than direct advertising: Digiday reports that fees paid to cloud providers reselling Claude accounted for $2.7 billion of the increase, and SaaStr's breakdown lists Accenture, Deloitte, KPMG, PwC, Salesforce and Snowflake among the partners Anthropic is paying to help sell its product into enterprises.

Brand as a disclosed risk, not just a budget line

What makes this filing worth a marketer's attention isn't only the dollar figure, it's that Anthropic chose to describe brand-building as a financial risk in a legal document reviewed by regulators and investors. The filing warns that keeping up the brand investment will keep costing real money, that falling behind better-known rivals is a genuine business risk, and that the company could get lumped in with other AI players' controversies regardless of its own conduct — language clearly aimed at distancing Anthropic from incidents at rival labs.

"This document makes it exceptionally clear that they consider their brand one of their greatest assets," said Nate Elliott, principal analyst for AI in marketing and commerce at eMarketer, who told Digiday the word "brand" appears more than 50 times in the risk-factors section alone.

The filing also leans on a specific point of differentiation, telling investors its monetization approach differs from unnamed competitors'. Digiday's reporting connects that line to OpenAI, which began running ads inside ChatGPT in February 2026 after Anthropic ran a Super Bowl spot promising Claude would stay ad-free. Whether or not that distinction holds up over time, Anthropic is treating its ad-free positioning as something worth defending in writing to investors, not just in marketing copy.

What this means if you market software

Most software marketers don't have an Anthropic-sized budget, and that's the point: the labs setting category expectations for "AI-native" branding now operate on a cost structure most SaaS companies can't match, funded by revenue growth most SaaS companies also can't match. If a buyer's mental model of how an AI product should look, sound and launch is being shaped by nine-figure PR budgets, a smaller vendor's quieter cadence will read as absence rather than restraint. That's a reason to be more deliberate about what you do spend on, not a reason to try to outspend anyone.

The partner-fee detail matters just as much as the headline multiple. A large share of Anthropic's "marketing" spend isn't advertising at all, it's channel economics: paying cloud marketplaces and systems integrators to put Claude in front of enterprise buyers. Marketers who treat marketplace listings, co-sell programs and SI relationships as a distribution afterthought now have a data point on what a frontier competitor pays to make that channel work.

The brand-as-risk-factor framing is also a useful internal argument. Marketers routinely struggle to justify brand spend to finance and the board as anything other than a nice-to-have. Here is a company about to go public stating, under legal scrutiny, that losing the brand fight is a material risk to the business. That's a stronger citation in a budget conversation than most internal brand decks will ever produce.

• Benchmark your own sales-and-marketing ratio against the category's growth-stage norm, not against Anthropic's absolute dollars — the 19%-of-revenue figure is the more transportable number
• Audit how much of your "marketing" spend is actually channel and partner enablement, and whether that split matches where your buyers actually discover you
• If you compete on an against-the-grain positioning claim, like ad-free or privacy-first, write down why it's defensible, the way Anthropic's filing does for its monetization stance