Simile Raises $200M at a $2B Valuation. It Wants to Simulate All 8 Billion Humans.

News Breakdown · FiscEdge Academy

Simile just closed a $200 million Series B at a $2 billion valuation, led by Greenoaks with Index Ventures, Bain Capital Ventures, A*, Factory, CVS Health Ventures and new investor Definition joining in. That values the company at roughly 20 times what it was worth just five months ago, when it emerged from stealth with a $100 million Series A also led by Index Ventures. In the time between those two rounds, Simile says revenue is up fivefold, headcount has passed 50, and the company has run tens of millions of simulations for Fortune 100 clients.

The number is the least interesting part. What Simile is actually selling is a substitute for your customers, and enterprises are already paying for it.

A foundation model trained to be a crowd

Simile builds what it calls "agentic twins": AI models trained to predict how real people will act, not just what they'll say, before a company ships a product, runs an ad, or changes a price. CEO Joon Sung Park is the Stanford researcher behind "Smallville," the 2023 experiment where generative AI agents lived out simulated daily lives, complete with parties and gossip. He co-founded Simile with Stanford professors Michael Bernstein and Percy Liang, and the company's stated goal is not subtle: build a foundation model that can simulate all 8 billion humans on Earth well enough that enterprises test decisions on the simulation before touching a real customer.

That is a direct pitch against an entire industry. Market research, focus groups, ad testing and pre-launch user studies are slow, expensive and often wrong by the time results come back. Simile is arguing that a well-trained behavior model can replace a meaningful chunk of that spend, and Fortune 100 healthcare, financial services and media companies are apparently buying the argument at scale.

Why the money moved this fast

A 20x valuation jump in five months is the kind of number that should make any founder suspicious, and it should. But look at what backed it up: this wasn't a follow-on round on a promise, it was a follow-on round on revenue that grew 5x and a customer list that expanded to "tens of millions of simulations" run in production. Index Ventures didn't just return, it stayed in as the company's numbers compounded. That is a very different signal than a hype-driven markup.

It also says something about where enterprise AI budgets are actually going right now. Boardrooms that were cautious about chatbots two years ago are comfortable paying for a system that claims to predict how their own customers will behave, because the ROI case, fewer failed launches, faster iteration, cheaper research, is easy to explain to a CFO. That is a more durable kind of AI demand than a general-purpose assistant subscription, and investors are pricing it that way.

What this changes for founders and operators

• Vertical, outcome-specific AI is still out-raising horizontal AI. Simile doesn't sell "an AI platform." It sells a specific, measurable replacement for market research. If you're building something "for AI" broadly, this round is another data point that founders solving one expensive, well-understood problem deeply are raising faster and at higher multiples than general-purpose plays.
• Revenue growth, not just usage growth, is what unlocks these markups. Simile's pitch to Series B investors leaned on 5x revenue growth in five months, not user count or engagement. If you're raising, the metric that moves a term sheet at this stage is money customers are already paying, not a roadmap of what they might pay for.
• "Test before you build" is becoming a fundable category on its own. For SaaS founders, the interesting read isn't just Simile's valuation, it's that enterprises will now pay to simulate a decision before committing real budget to it. If your product touches pricing, onboarding, or messaging, expect customers to start asking whether you can show simulated outcomes before they commit.
• A returning lead investor is a stronger signal than a new logo. When evaluating your own comps or a competitor's raise, weight it more heavily if existing investors are doubling down than if a brand-name fund is showing up cold. Conviction from people who already have the data room is worth more than a new name on the cap table.

If you remember one thing

Simile didn't get a 20x markup in five months for a good demo, it got one because revenue grew 5x and its existing investor stayed in to write a bigger check on real numbers. If you're building anything AI-labeled and fundraising in this market, the lesson isn't "move fast," it's "have the growth curve ready before you ask for the number." Valuation follows proof, not the other way around.

---

We teach how to read a cap table and a growth curve like this in FiscEdge's financial modeling track, and how to position a vertical AI product for a raise in AI for entrepreneurs. If you're building the kind of metrics-first product that raises rounds like this one, building SaaS with AI and our breakdown on what actually drives unit economics are the right next reads. Browse the full blog for more breakdowns like this. Follow @fiscedge for daily Business & AI analysis.