ICONIQ Capital just dropped their annual State of Software report, and inside the 73 (!) pages are some data that shows just how much the software industry is undergoing a fundamental transformation.
Here are my top 5 take-aways (and 5 more bonus ones):
1. AI-Native Companies Are Hitting $100M ARR with Teams of Less Than 100-150 People
Traditional SaaS wisdom says you need hundreds of employees to reach $100M in ARR. AI-native companies are throwing that playbook out the window.
The numbers: Cursor reached $100M ARR in one year with approximately 19 employees. Lovable hit the milestone in 8 months with 45 people. ElevenLabs got there in roughly 2 years with around 150 employees.
At a practical level, they often have tiny GTM teams to start, and often for longer, and fairly tight dev teams. Perplexity got to 5,000 customers with just 5 folks on the sales team:
Compare that to traditional SaaS benchmarks where top-quartile companies took 5+ years and 500-700+ employees to reach $100M ARR. We’re talking about 5-10x higher revenue per employee.
This isn’t just about better unit economics—the burn rates are often still substantial. But in many ways, it’s a completely different business model. Many AI-native levels run PLG models to an extreme, relying on strong viral demand and near instant ROI. And fewer customer success reps, smaller sales teams, leaner ops organizations.
2. AI-Native Companies Reach $100M ARR 2-3x Faster Than Top Quartile SaaS
We all know this, but the time compression here is staggering. AI-native companies are reaching $100M ARR in 4-8 quarters. Top quartile traditional SaaS companies take 18-20 quarters to hit the same milestone.
That’s not incremental improvement—it’s a completely different growth curve.
Look at the trajectory: Cursor and Lovable hit $100M ARR in under a year. Traditional top-performing SaaS companies like those in ICONIQ’s benchmark took 5+ years to get there. Even the fastest-growing SaaS companies of the 2010s couldn’t match this pace.
This isn’t just about better product-market fit or more efficient go-to-market. The underlying economics are fundamentally different. When your product can onboard customers instantly, demonstrate value in days instead of months, and scale without proportional headcount growth, you compress years of traditional SaaS growth into quarters.
The implication: we’ll all begun to completely rethink what “fast growth” means. A company growing 100% year-over-year used to be exceptional. AI-native companies are growing 200-300% year-over-year and treating that as table stakes.
3. AI Companies Burn More Cash — But Also Have Better Capital Efficiency. It’s Not a Paradox But … It’s Complicated.
This one breaks your brain at first: AI-native companies under $100M ARR have a median FCF margin of -126% (they’re burning cash at 126% of revenue). That’s more than double the -56% for non-AI companies.
Yet their burn multiple—the key metric for capital efficiency—is actually better: 0.4x versus 1.8x for non-AI companies.
Translation: AI companies are burning more absolute dollars, but they’re generating new ARR so much faster that each dollar burned produces more revenue growth. They can afford to burn harder because the growth rates are exponential rather than linear.
This suggests we need entirely different frameworks for evaluating AI-native businesses. Traditional SaaS metrics around payback periods and magic numbers may not capture what’s really happening when you can scale revenue 3x faster than historical benchmarks.
4. GTM for AI Products Is Flipped: 55% of High-Growth Teams Are in Post-Sales
Traditional SaaS companies put 55% of their GTM headcount in sales roles. For high-growth AI-native companies, that ratio is flipped: 47% in sales, but 31% in post-sales (versus just 23% for traditional SaaS).
The rise of forward-deployed engineers tells the story. When your product is technically complex and requires deep integration, you need engineering talent embedded with customers to drive adoption and expansion. The “land and expand” motion for AI products is fundamentally more technical than traditional software.
Job posting data confirms the shift: Forward-deployed engineer job postings have increased from around 30 per month in early 2024 to 375 by April 2025—a 12x increase.
This isn’t just a staffing decision. It reflects a deeper truth: AI products require a different GTM motion. Traditional sales skills matter less than technical implementation expertise. The sale isn’t complete until the AI is actually producing measurable business outcomes, which requires ongoing technical engagement.
5. Private AI Funding in First Half of 2025 Already Exceeds All of 2024
Private AI/ML companies raised $377 billion in the first half of 2025. That’s more than the $363 billion raised in all of 2024.
The average deal size jumped from $17.2M in 2024 to $35.9M in 1H 2025—more than doubling.
This isn’t just froth. When you dig into the deployment patterns, enterprise adoption of AI products is accelerating significantly:
- 80% of companies report active AI experimentation or implementation
- 94% of public software company earnings calls mention AI
- High AI adopters report 38% average productivity gains in R&D
The capital is flowing toward real revenue and real adoption, not just demos and POCs. We’re past the “AI winter” concerns and deep into a genuine platform shift.
5 More Stats That Didn’t Make the Top 5 Cut (But Are Still Super Interesting)
- Offshore headcount jumped from 24% to 30% YoY, with most growth in engineering and support roles. Combined with AI tooling, companies are finding 40-50% cost arbitrage opportunities that are finally moving the needle on burn multiples.
- Forward-deployed engineer job postings increased 12x in one year, fastest growing in software—from ~30/month in early 2024 to 375/month by April 2025. The fastest-growing job title in software isn’t “AI engineer,” it’s customer-facing technical roles.
- Top quartile companies maintain 1.5x+ net magic number regardless of scale. Even $500M+ ARR companies are generating $1.50+ in net new ARR for every dollar of S&M spend in the prior quarter. This metric has stopped declining for the first time since 2021.
- ARR per FTE is growing faster than OpEx per FTE across all company stages. Median ARR per FTE has increased from $182K to $237K over the past 5 years, while OpEx per FTE has remained relatively flat at $220-230K. The labor productivity curve is finally bending.
- 94% of public software earnings calls now mention AI, up from essentially 0% in Q2 2022. More tellingly, mentions of “AI agents” have spiked in recent quarters—suggesting the conversation is moving from “AI features” to “AI doing actual work.”
What It All Means
These stats all point to the same conclusion: we’re not just seeing an incremental improvement in software economics. We’re witnessing the emergence of a fundamentally different category of software business. One where churn and retention rates may in some case be a work in progress, and where token costs may be material. But where growth and market pull are at a scale we’ve never seen before.
AI-native companies can scale faster, with fewer people, using different GTM motions, while burning capital more efficiently than traditional SaaS ever could. The old playbooks—the ones written during the 2010s-2022 SaaS boom—increasingly only partially apply.
The companies that win over the next decade won’t just be the ones that add “AI features” to existing products. They’ll be the ones that rebuild from first principles around what’s possible when intelligence is a commodity and humans focus on the 20% of work that truly requires judgment.
The metrics in ICONIQ’s report aren’t just interesting data points. They’re early signals of a new software paradigm taking shape in real-time.
Data sourced from ICONIQ Capital’s State of Software 2025 report, analyzing 127 software companies including portfolio companies and select public companies as of Q2 2025.
