Sarah
Wang

Before the AI funding frenzy made "growth-stage enterprise investor" sound like a lottery ticket, Sarah Wang spent years doing the unglamorous version: studying enterprise software metrics, sitting on ZoomInfo's board at TA Associates, and building a mental model of what actually makes a B2B company scale. Most investors have one mode - either venture (conviction on what could be) or growth (rigor on what is). Wang has both simultaneously, which is why David George, her hiring manager at a16z, called her "as good as it gets."

She joined Andreessen Horowitz in 2019. Within three years she was General Partner - promoted in January 2022. By then her portfolio was already forming the shape it has today: Sourcegraph, Crossbeam, Wiz, Hex. She was building conviction in enterprise infrastructure before the mainstream had finished debating whether remote work was permanent.

Then came the AI wave, and Wang was positioned precisely at its intersection with enterprise spending. She led a16z's investments in Cursor (the AI-native code editor that redefined developer productivity), Gamma (PowerPoint's AI replacement that hit product-market fit faster than almost anything in its cohort), Decagon (enterprise AI agents), Thinking Machines Lab, and Character.ai. She also works closely with a16z's growth positions in OpenAI, SSI, Databricks, SpaceX, Samsara, and Figma - a portfolio that reads less like a bet and more like a who's who of the companies that are actually winning.

Martin Casado, her colleague at a16z, publicly called her "the best growth investor in the entire industry." That kind of peer recognition in a firm of that caliber means something specific: it means she's the person they put on the hardest problems.

The frontier model investment thesis she's articulated is one of the more precise frameworks to emerge from the current AI moment. Traditional venture and traditional growth investing have different questions: venture asks "what if this works?" and growth asks "is this working?" Frontier model companies, Wang has observed, require both at once - they need growth-scale capital from day one while still being early-stage founder bets. That's a new kind of deal that requires a new kind of investor.

Her read on the talent market is equally sharp. "You could be an L5 and get an offer in the tens of millions," she noted, describing how AI compensation spirals don't just affect headline acqui-hires - they reshape entire salary bands and startup economics across the industry. She's seen the fishbowl effect up close: the intense media scrutiny that creates anxiety loops for AI founders who can't tell what's real signal and what's Twitter noise. Her frustration with that gap is pointed: "I have never lost so much faith in the accounts on Twitter - couldn't be further from the truth."

On specialization versus generalization in AI: "There's clear value to specialization - they're still fricking number one, right?" She's not a zero-sum thinker. The portfolio shows it - she backs both foundation model companies and the application layer, both infrastructure and consumer AI, because she's read the data on where enterprise spending is actually going.

Wang's annual enterprise AI survey - now in its third year, covering 100 Global 2000 CIOs - has become one of the most-read data sets in Silicon Valley. Not because it flatters anyone's narrative, but because it's methodical in the way that only someone with genuine enterprise pattern recognition can pull off. When she wrote that "reports of the app apocalypse are greatly exaggerated," she wasn't cheerleading - she had the data showing third-party application adoption continuing to grow across use cases despite the foundation model boom.

The survey findings move markets. When Anthropic posted a 25% increase in enterprise penetration since May 2025 - leapfrogging expectations and reaching 44% production use - that signal came from Wang's research. When she documented that 54% of enterprise respondents said reasoning models accelerated LLM adoption (faster time to value, better trust through accuracy), it gave portfolio companies and enterprise CIOs alike a shared vocabulary.

Her written work on enterprise software - "Anatomy of an Enterprise Platform Company," the growth metrics guide, the pricing and packaging frameworks for B2B AI - reflects the same sensibility. These aren't blog posts. They're operating manuals. And they're widely used by founders because they're grounded in what Wang has seen actually work, at scale, in real enterprise accounts.