Used to brag about our team size. "We're at 14 people now!" felt like progress.
Then we laid off 9 people in 2023. Most painful moment of my career.
Two years later, we're still 5 people. Revenue is higher. Product is better. Customers are happier.
I stopped celebrating headcount because I finally understood what it actually measures.
The Headcount Status Game
Every founder plays it:
"We just hit 10 employees!" "Scaling to 25 people this quarter!" "Series A means we can finally build a real team!"
The underlying assumption: More people = more success = more important company.
I believed this completely. Posted LinkedIn updates about new hires. Felt proud introducing our "team of 14" to investors.
The headcount felt like validation that we were building something real.
Then AI changed the math. And a recession forced us to confront reality.
The Historical Logic That Made Sense
For most of business history, headcount WAS capacity:
Fast food restaurants: Need bodies to operate fryers, take orders, clean tables. No fryer operates itself.
Airlines: Need pilots, crew, ground staff for safety. Physics and regulations demand human presence.
Factories: Need workers on assembly lines, quality control, machine operation. Production requires people.
Professional services: Need consultants, analysts, project managers. Billable hours scale with people.
The equation was simple: More people = more capacity = more output = more revenue.
This worked because humans were the execution layer. There was no alternative.
Software companies fell into the same pattern because software development was labor-intensive:
- More developers = more features
- More salespeople = more deals
- More support staff = more customers served
- More product managers = better coordination
We hired like a factory because we thought like a factory.
When AI Broke the Math
The shift happened gradually, then all at once:
2021: AI tools were novelties. GitHub Copilot was interesting but not essential.
2022: ChatGPT launched. Useful for content but not core business functions.
2023: Claude, GPT-4, Cursor. AI became capable enough to replace entire job functions.
2024: AI tools matured. What used to require full-time employees now takes part-time AI assistance.
The new equation:
- 3 people with AI can outproduce 15 people without it
- Execution tasks get automated
- Coordination overhead becomes the biggest cost
- Headcount transforms from capacity to liability
My personal wake-up call came during our 2023 budget review.
The 2023 Layoff: The Most Painful Lesson
The setup:
14 people on the team by Q4 2022. Everyone had been there 1-2 years. Real relationships, not just professional ones.
The reality check:
Revenue growth slowing. Runway shrinking. Market conditions deteriorating.
The analysis:
Looking at actual work being done vs. headcount:
- 3 people doing work that directly impacted customer value
- 6 people doing work that could be replaced by AI tools
- 5 people doing work that probably didn't need to exist
The decision:
Cut down to 5 people. Keep only the roles that AI genuinely couldn't replace.
The execution:
One of the most painful moments of my career. Small company layoffs hit different than big company layoffs.
These weren't faceless employees. These were people I'd worked closely with for years. I knew their career goals, their families, their contributions.
The guilt was overwhelming.
What Happened Next
The surprising aftermath:
Could have hired back. Chose not to.
Over the next 24 months:
- Revenue increased 47%
- Customer satisfaction improved (94% vs 87% previously)
- Product shipped faster (2-week cycles vs 6-week cycles)
- Team stress decreased (less coordination overhead)
- Profit margins improved 340%
Used AI for everything we thought we might need a person for:
Content creation: Claude writes better copy than our previous marketing hire Customer support: AI handles 80% of inquiries, human handles 20% Data analysis: ChatGPT analyzes metrics faster than our previous analyst Design work: AI + design tools replace junior designer role QA testing: Automated testing + AI covers most QA tasks
The uncomfortable truth: We were more productive with fewer people.
What AI Replaced vs What It Couldn't
AI successfully replaced:
Execution tasks we would have hired for:
- Content writing and copywriting
- Basic data analysis and reporting
- Customer support for common issues
- Social media management
- Meeting summaries and documentation
- Basic design and graphic work
- Code review and debugging assistance
Total estimated replacement value: ~4 full-time equivalent roles
AI couldn't replace:
Strategic thinking and direction:
- Market positioning decisions
- Product roadmap prioritization
- Customer relationship management
- Crisis decision-making under uncertainty
- Creative problem-solving for novel challenges
- Cross-functional collaboration and alignment
The nuance most people miss:
I focused on the cost savings. I didn't count the strategic thinking we lost.
The people we laid off weren't just executing tasks. Some could have reshaped the company's direction.
The counterfactual is invisible: What would we have built with more diverse thinking?
The Hidden Costs of the Lean Team
What I gained with 5 people + AI:
- Lower overhead and higher margins
- Faster decision-making (no committee paralysis)
- Clearer accountability and ownership
- Reduced coordination complexity
- Higher productivity per person
What I lost:
- Diverse perspectives on product direction
- Specialized expertise in areas I'm weak
- Fresh thinking that challenges my assumptions
- Capacity to explore multiple opportunities simultaneously
- Buffer for handling unexpected challenges or opportunities
The trap:
- Easy to measure cost savings from fewer people
- Hard to measure lost innovation from groupthink
- Echo chambers execute faster but explore less
Example of what we missed:
In 2023, AI agent workflow automation became a major market opportunity. Our remaining team focused on core product improvements.
With our previous team structure, we had two people who were exploring adjacent opportunities. They likely would have identified and built AI agent features 6 months earlier.
That missed timing potentially cost us $200K+ in additional revenue.
But I only see this in hindsight. At the time, staying lean felt obviously correct.
The Productivity Paradox
The counterintuitive reality:
Individual productivity with AI: 5-10x improvement Team productivity with fewer people: 2-3x improvement Innovation speed with smaller team: Often slower
Why the gap?
Coordination overhead scales with team size squared:
- 5 people = 10 potential communication paths
- 10 people = 45 potential communication paths
- 15 people = 105 potential communication paths
But innovation often requires diverse perspectives:
- Different backgrounds bring different solution approaches
- Specialized expertise catches problems generalists miss
- Fresh eyes question assumptions that become invisible to core team
- Multiple work streams allow parallel exploration
The AI multiplier helps with execution but not with strategic diversity.
What Headcount Actually Measures
What I thought headcount measured:
- Company growth and success
- Market validation and investor confidence
- Capacity to tackle bigger challenges
- Progress toward becoming a "real" company
What headcount actually measures:
- Coordination overhead and communication complexity
- Cash burn rate and runway reduction
- Commitment to specific organizational structure
- Historical decisions about work distribution
What headcount doesn't measure:
- Actual productive output or value creation
- Quality of decision-making or strategic thinking
- Customer satisfaction or market impact
- Adaptability to changing market conditions
- Profitability or business sustainability
The realization: I was celebrating an input (people) rather than outputs (value created).**
The New Status Game
What I celebrate now:
Revenue per employee:
- 2022: $47K revenue per person (14 people, $660K revenue)
- 2024: $168K revenue per person (5 people, $840K revenue)
- 257% improvement in productivity
Customer satisfaction metrics:
- Net Promoter Score increased from 67 to 89
- Support response time improved 340%
- Feature request implementation time cut in half
Profit margins:
- 2022: 23% profit margin (high headcount overhead)
- 2024: 78% profit margin (lean team + AI efficiency)
Decision-making speed:
- Product decisions: 2 weeks average (vs 6 weeks previously)
- Strategic pivots: 1 month (vs 3-4 months previously)
- Customer issue resolution: Same day (vs 3-5 days previously)
The new equation: Output per person matters more than total output.
Industry Patterns I'm Seeing
Companies celebrating headcount reduction:
Basecamp: Cut from 57 to 32 people, revenue stayed flat, profit increased 85% Buffer: Reduced team 25%, maintained growth rate, improved work-life balance GitLab: Stayed lean while scaling to IPO, highest revenue per employee in sector
Companies still hiring like it's 2019:
Many SaaS startups: Adding people for functions AI could handle Enterprise software: Building bloated teams for "enterprise readiness" Venture-backed companies: Hiring because investors expect headcount growth
Want the full playbook? I wrote a free 350+ page book on building without VC.
Read the free book·Online, free
The pattern: Companies optimizing for AI leverage outperform those optimizing for headcount.**
The Optimal Team Size Question
What I learned about team composition:
The 5-person core:
- 1 technical leader (me)
- 1 product/customer specialist
- 1 operations/business specialist
- 1 designer/UX specialist
- 1 senior developer
Plus AI for:
- Content creation and marketing
- Customer support and documentation
- Data analysis and reporting
- Quality assurance and testing
- Administrative and operational tasks
This structure provides:
- Diverse perspectives (5 different backgrounds)
- Specialized expertise in key areas
- Low coordination overhead
- High individual accountability
- AI augmentation for execution tasks
When we consider hiring person #6:
Questions I ask now:
- What type of thinking would this person add?
- Could AI + existing team handle this function?
- Will coordination overhead offset productivity gains?
- Does this create opportunities for innovation we're missing?
- Is this hire about capability or capacity?
The test: If the hire is about doing more of what we already do, AI probably handles it. If it's about thinking differently about what we should do, consider the hire.
The Contrarian Take on AI + Teams
What everyone says: "AI will make every team more productive."
What I learned: "AI makes small, high-quality teams incredibly productive but can amplify groupthink if you're not careful."
The nuance:
AI multiplies existing team capabilities:
- Good teams become great with AI assistance
- Dysfunctional teams become dysfunctional faster with AI
- Diverse teams get better ideas; homogeneous teams get faster execution of similar ideas
The risk:
- Lean team + AI = incredible efficiency
- Lean team + AI - diverse perspectives = efficient execution of potentially wrong strategy
The solution:
- Keep team small but maximize cognitive diversity
- Use AI for execution, humans for strategic thinking
- Bring in external perspectives through advisors, customers, consultants
- Resist hiring for capacity, hire for different types of thinking
What This Means for Different People
For early-stage founders:
- Don't hire until you've exhausted AI + existing team capabilities
- When you do hire, optimize for cognitive diversity over specialized execution
- Measure productivity per person, not absolute productivity
- Celebrate lean team achievements rather than headcount milestones
For investors:
- Revenue per employee becoming more important metric than total headcount
- Portfolio companies with lean teams + AI leverage will have better unit economics
- Headcount growth plans should justify why humans beat AI for specific functions
- Look for teams that can articulate their human-vs-AI strategy
For job seekers:
- Execution-only roles increasingly automated; strategic thinking roles safer
- Demonstrate ability to work with AI tools, not compete against them
- Focus on uniquely human capabilities: creativity, empathy, strategic thinking
- Small teams offer more diverse experience but less job security
For managers:
- Team coordination skills becoming more valuable than team building skills
- Managing AI-augmented humans requires different approach than managing humans alone
- Focus on output quality and strategic direction rather than task completion
- Small teams require higher hiring standards and better onboarding
The Uncomfortable Questions
For founders obsessing over headcount:
Are you hiring because you need different thinking or just more capacity?
Could AI + your existing team handle this function better than a new hire?
Are you measuring inputs (people) or outputs (value created)?
Is team size a vanity metric that's masking underlying business issues?
For investors pushing headcount growth:
Does adding people actually increase company value or just burn rate?
Are you conflating team size with company maturity and scale?
How do you evaluate companies that achieve more with smaller teams?
What happens to portfolio company economics when AI replaces their planned hires?
For teams celebrating growth:
Is your team genuinely more capable or just more expensive?
Are you adding coordination complexity faster than productive capacity?
Could you achieve the same goals with AI augmentation instead of new hires?
Are you hiring to solve problems or to feel like you're scaling?
Action Steps for Founders
This week:
- Calculate your current revenue per employee
- List all functions new hires would perform
- Test AI alternatives for each function before hiring
- Identify what type of thinking your team is missing
This month:
- Audit your existing team for AI augmentation opportunities
- Set productivity metrics that reward output over input
- Stop celebrating headcount milestones in team communications
- Start tracking decision-making speed and quality metrics
This quarter:
- Implement AI tools for routine execution tasks
- Create hiring criteria focused on cognitive diversity
- Build advisor network to supplement internal perspectives
- Optimize team structure for AI-augmented productivity
This year:
- Prove you can scale revenue without proportional headcount growth
- Build competitive advantage through lean team + AI leverage
- Help other founders understand headcount vs productivity trade-offs
- Create playbook for AI-augmented team optimization
Why This Matters
Because the startup world still measures success with outdated metrics.
Because AI fundamentally changed the relationship between people and productive capacity.
Because small teams with AI leverage will outcompete large teams without it.
Because celebrating headcount encourages inefficient scaling and capital allocation.
The old game: Hire fast, scale big, hope for exit before unit economics matter.
The new game: Stay lean, leverage AI, optimize for output per person, build sustainably profitable business.