AI Startups are Dominating Traditional Software in one Key Metric

We’re on a path to seeing a solo-employee startup reach a $1B valuation, and companies generating over $100M in revenue per employee. This week, Saastr conference in San Mateo is happening, and we should compare and contrast the old way (Trad-SaaS) vs new (AI-SaaS).

Using my background as a former Industry Analyst, let’s review history and then do a current market analysis comparison of Trad Saas vs AI SaaS

Technology Eras:
I’ve been in tech through several major waves. Here’s what I’ve experienced:

• DotCom Era

  • Sky-high valuations with few having business models

  • Lots of employees.

• Web2 Era

  • Valuations driven by user adoption, but revenue models came years later.

  • Lots of employees.

• Sharing / On-Demand Economy Era

  • Real revenue, but reliant on contractors and marketplaces.

  • Lots of employees and contractors/affiliates

• GenAI Era

  • Real revenue.

  • Ultra-lean teams. In fact, many “employees” are AI agents.

These new AI-native startups follow “AI-First” principles: use AI before hiring humans, and maximize efficiency from day one. In fact, many of the “employees” are AI Agents —not humans. I was recently quoted in Fast Company on this very topic: AI coding tools could bring us the ‘one-employee unicorn’

Comparison: Revenue Per Employee (RPE) In Thousands of Dollars

AI Productivity Metric: Revenue Per Employee
Revenue Per Employee (RPE) over an annual rate, is a key metric for evaluating a company’s productivity—and in the age of AI, it’s more important than ever. You can contrast traditional companies that are only now adopting GenAI with the newest AI-native startups. These startups not only build AI products but embed AI into their culture from day one, treating it as the first resource —before hiring additional humans.

Traditional: Trad-SaaS hires thousands of employees:
Note: It’s challenging to get accurate data in this ratio.

• $200K: Average traditional SaaS company, Revenue Per Employee (RPE)

• $318K: HubSpot

• $479K: Salesforce

• $700K: Adobe

• $200K: Box

vs

New AI Startups “AI-SaaS”, dozens to hundreds of employees:

1. $12,500K: Midjourney, that’s $12.5M Revenue Per Employee (RPE)

2. $5,000K: Anysphere (who makes Cursor)

3. $3,000K: Cal AI

4. $2,500K: Mercor

5. $2,500K: Chai Research

6. $2,000K: Eleven Labs

7. $2,000K: Stackblitz

8. $1,818K: Fal AI

9. $1,786K: Gamma

10. $1,667K: Lovable

Analysis of Top 10 Lean AI Native Startups:

• A mere 24 employees on average across those ten AI-native startups.

• The top 10 AI companies have an average Revenue Per Employee (RPE) of $3.48M—significantly higher than the traditional SaaS average of $200K.

• Even with the outlier removed (Midjourney), the average RPE for companies #2 through #10 is still $2.47M—an order of magnitude greater than the traditional SaaS benchmark of $200K RPE.

Trad SaaS CEOs to employees: “We must shift to be an AI-first company”
We're seeing a clear signal from tech leadership: AI isn't a side project—it's the core strategy. CEOs from Box, Shopify, Duolingo, and others have issued public mandates to make their companies “AI-First.” This means more than adopting tools—it’s a cultural and operational shift. Teams are expected to rethink workflows, products, and even org design around AI capabilities. The message is simple: adapt fast, or fall behind.

I’m investing in AI-First Startups
At Blitzscaling Ventures, where I’m deploying AI investments, we backed a company that has about 25 employees but 150 AI agents and they are rapidly growing in adoption and revenue.

Learn more about the Lean AI Native Companies
Check out to the Top Lean AI Native Companies Leaderboard. Thanks to Henry Shi for curating it.

Llama Lounge, is Two-Years Old
I run Llama Lounge, an AI Startup Event Series (for AI Founders, Investors and Corporate AI leaders), the next event is at Zoom HQ, on May 22nd. Llama Lounge just hit our two year aniv, see before and after photo.

Hey, what about Apple, Microsoft, Google, Nvidia, or Amazon? That would be an apples-to-oranges comparison, as these companies have hardware businesses, sometimes operate retail divisions, rely on low-cost contractors for hardware production, and often obscure true headcount through extensive use of contracting firms. For now, we’re keeping the comparison SaaS to SaaS.

Questions for you:

1. Will the AI-SaaS startups hire more humans?

2. If yes, will they hire as many humans as Trad-SaaS?

3. Will Trad-SaaS companies hire more or fewer employees?

4. Will Trad-SaaS companies increase investment in AI?