The idea that a leading growth investor would downshift financial metrics and elevate founder judgment to the center of the investment process feels counterintuitive—especially at scale. Yet that is exactly what Ethan Choi, known for backing companies like Ramp, ClickHouse, Vercel, Glean, and Bridge, has done. After years of weighting his decisions roughly 80% toward financial performance and 20% toward founder evaluation, Choi has inverted the formula. Founder adaptability now represents 90% of the decision. Metrics have been relegated to supporting evidence.
This pivot is less personal preference and more structural signal. AI has fundamentally accelerated the operating tempo of software companies. Markets now move at what Choi describes as “Ironman suit” productivity levels—cycles that once played out over years now compress into months. In this environment, historical numbers lose their predictive power. What remains stable is the founder’s ability to navigate discontinuity. That is the core of the new calculus.
AI’s velocity is challenging the foundation of traditional growth‑stage evaluation. Over the past decade, markets have absorbed the equivalent of half a century of technological transformation. For investors, this compression renders backward‑looking signals increasingly unreliable. Metrics built on historical growth, cohort stability, or sales efficiency struggle to forecast outcomes when product velocity, customer expectations, and competitive dynamics shift simultaneously.
The valuation framework is evolving accordingly. Revenue multiples—once the anchor of software pricing—are giving way to free cash flow and price‑to‑earnings ratios tied to AI‑native revenue streams. Investors are no longer willing to apply premium multiples to legacy software revenue that may be eroded or replaced entirely by AI‑automated workflows. The market now separates companies capable of converting AI into durable economics from those reliant on legacy seat‑based licensing.
Some business models are particularly exposed. Lightweight horizontal SaaS players face intense margin compression as AI narrows the moat between tools. Mid‑market platforms that cannot attract high‑end AI engineering talent also risk stagnation. The experience of companies like Intercom and Airtable illustrates the pressure: both were forced into accelerated reinvention cycles, not due to mismanagement but because the baseline for product expectations shifted underneath them. These stories are becoming less exception than preview.
Choi’s shift also sheds light on a broader movement: the erosion of traditional stage boundaries. If founder quality is the single most persistent differentiator, the distinction between seed, Series A, or growth becomes less meaningful. Capital naturally migrates earlier, where pricing is more flexible and investors can build conviction around the one constant variable—the founder’s capacity to execute amid volatility.
This philosophy introduces new criteria for assessing talent. Beyond functional excellence or managerial track record, the bar now includes whether executives have “IC’d themselves”—demonstrated the ability to operate as hands‑on individual contributors when speed demands it. AI tools further amplify this expectation, making it feasible for leaders to personally drive product, analysis, or GTM experimentation at levels once impossible.
The DualEntry/NetSuite dynamic illustrates how this new era unlocks disruption. Historically, migrating away from a dominant financial system required armies of consultants and extensive enterprise coordination, effectively freezing market share. AI‑assisted migration changes the feasibility equation. Emerging players like DualEntry can now automate transitions that were once prohibitive, creating genuine openings in markets long considered sealed. For investors, such shifts redefine what counts as an addressable opportunity.
As capital allocators recalibrate their frameworks, several principles stand out. First, founder adaptability outweighs static assessments of product‑market fit or revenue trajectory. A strong product today offers little assurance in a landscape where AI competitors can replicate functionality in weeks. The founder’s ability to reposition the company, reorganize teams, and incorporate new capabilities becomes the core predictor of durability.
Second, revenue architecture matters. Investors must determine whether a business is moving toward inference‑ or usage‑based models aligned with AI value creation, or whether it remains tied to seat‑based licensing that may quietly erode. Third, leadership should demonstrate comfort operating as AI‑augmented individual contributors, not solely managers. In a compressed cycle, the most effective executives personally build, iterate, and validate assumptions.
Finally, investors must distinguish between AI‑native companies and those attempting retrofit. Retrofitting can work, but only if the organization has the talent density, capital, and cultural willingness to rebuild core workflows. Many companies underestimate the required commitment.
This framework is directional, not final. As AI compounds, the investment rulebook will continue to evolve. For now, the shift toward founder‑first conviction reflects a pragmatic response to the most volatile technological cycle in decades—one in which adaptability, not historical performance, defines the frontier of venture success.
