AI Markets: Deep Dive with a16z's David George
This episode analyzes the state of AI in 2025, revealing that the best AI companies are growing 2.5x faster than non-AI companies while running significantly leaner operations—achieving $500K-$1M in ARR per employee versus the traditional $400K benchmark. The key insight: this unprecedented efficien
47mKey Takeaway
This episode analyzes the state of AI in 2025, revealing that the best AI companies are growing 2.5x faster than non-AI companies while running significantly leaner operations—achieving $500K-$1M in ARR per employee versus the traditional $400K benchmark. The key insight: this unprecedented efficiency isn't primarily from AI tools yet, but from extraordinary product-market fit creating such strong demand that companies need fewer resources to scale. The critical action for all companies, pre-AI or post-AI: "adapt or die" by reimagining products with AI at the core and implementing AI tools across every function, especially engineering where recent advances in coding assistants are creating 10-20x productivity gains.
Episode Overview
David Ulevitch from Andreessen Horowitz presents a comprehensive analysis of the AI market in 2025, examining both the demand and supply sides of the AI economy. The episode covers revenue growth acceleration in AI companies, their operational efficiency metrics, customer engagement data from portfolio companies, public market dynamics, and the massive infrastructure buildout underway. Key themes include the sustainability of AI company growth, the adaptation imperative for legacy companies, and early signs of how AI is transforming both products and internal operations.
Key Insights
AI Companies Growing 2.5x Faster with Better Unit Economics
The fastest-growing AI companies are reaching $100M in revenue significantly faster than the fastest SaaS companies in their era, but they're doing it while spending LESS on sales and marketing than their SaaS counterparts. This isn't about spending more to grow faster—it's about having products with such strong demand that they require fewer resources to scale. The top-performing AI companies are growing at 693% year-over-year.
ARR per Employee: The New Efficiency Metric
The best AI companies are achieving $500K-$1M in annual recurring revenue per full-time employee, compared to the traditional software benchmark of $400K. This metric captures total company efficiency—not just sales and marketing, but also R&D and overhead. While some of this comes from AI tooling, most currently stems from extraordinarily strong product-market fit requiring less go-to-market resources.
Gross Margins as a Badge of Honor
Lower gross margins for AI companies can actually be positive—indicating high usage of AI features and inference costs. If gross margins are too high in an AI pitch, it may signal that AI features aren't actually being used by customers. The expectation is that inference costs will decline over time, improving margins while usage stays high.
December 2024: The Coding Productivity Inflection Point
December marked a turning point in AI-assisted coding. One founder rebuilt a product from scratch with two AI-native engineers using unlimited access to Claude, Codex, and Cursor, achieving 10-20x faster progress than before. This has led him to conclude that his entire product and engineering organization needs to work this way within 12 months, fundamentally rethinking team design and where product management and design even fit into the process.
The 'Electricity vs. Blood' Decision Framework
A pre-AI company CEO now asks for every task: 'Can I do it with electricity or do I need to do it with blood?' This represents the extreme mindset shift required—treating human resources and AI agents as fundamentally different inputs, and defaulting to AI wherever possible. This framing helps companies systematically identify where AI can replace or augment human work.
Business Model Evolution Still Early—A Window for Incumbents
While technology and products are shifting rapidly, business model evolution is still in early days. The spectrum runs from licenses → seat-based subscriptions → consumption/usage-based → outcome-based pricing. The most disruptive scenario is when both technology AND business model shift simultaneously. Since outcome-based pricing is still nascent (mainly viable in customer support), pre-AI companies have time to adapt on the product side without facing dual disruption.
Notable Quotes
"AI demand side is crazy. The actual uptake growth quality of companies in AI is extremely encouraging from our standpoint. Companies are starting to run themselves better."
"Adapt or die. You need to adapt to the AI era or die. That's both on the front end and the back end."
"I now ask the question for every task that we now need to complete: can I do it with electricity or do I need to do it with blood?"
"He said he thinks it's going somewhere between 10 and 20x faster than progress that they had before. The bills that they have associated with that is actually high enough that it will cause him to rethink what his entire organization will look like."
"Low gross margins for AI companies are sort of a badge of honor in the sense that we want to see if low gross margins are a result of high inference costs. One, that means people are using AI features and two, we have a belief that those inference costs over time are going to come down."
"The leading tech companies are the best businesses in the history of the world. If you just look over a long period of time, they have shown margin improvement that suggests that is probably true."
Action Items
-
1
Audit Every Business Function with the 'Electricity vs. Blood' Framework
Go through each task and process in your organization and ask: 'Can this be done with electricity (AI) or does it require blood (human labor)?' Start with coding, customer support, and high-volume repetitive tasks. Give teams unlimited budgets on AI tools in pilot areas to discover what's possible.
-
2
Implement Unlimited AI Coding Tool Access Immediately
Roll out Claude, Cursor, GitHub Copilot, or similar tools to your entire engineering organization with unlimited usage budgets. Track productivity metrics and bill costs to understand the ROI. Based on early results showing 10-20x gains, this investment pays for itself many times over.
-
3
Measure and Optimize ARR per Employee as a Key Metric
Add ARR (or revenue) per FTE as a core metric in your business reviews. This captures total company efficiency across all functions, not just sales efficiency. Target $500K+ per employee for AI-era companies. Use this to identify where you're over-resourced or under-leveraging AI.
-
4
Assess Your AI Product Adaptation Urgency
If you're a pre-AI company, don't just bolt on chatbots. Reimagine your core product with AI at the center. Ask: 'If we started today, how would we build this with AI?' Then create a roadmap to either rebuild from scratch (like the example with two engineers) or systematically transform your existing product. You have a window before business models shift, but the product transformation must start now.