Why is scaling a B2B SaaS company from $1M to $10M ARR so hard?
This stage is the most dangerous because it requires transitioning from a startup to a scale-up, and the playbook that got you to $1M won't work at $5M or $10M. At $1M, the founder closes most deals, the product team builds on gut feel, and customer success means the CEO calling upset customers. None of that scales, and companies that stall at $2-3M are almost always the ones where the founder can't let go of those early habits.What is the most important metric when scaling B2B SaaS?
Net Revenue Retention above 100% is the single most important metric in the foundation stage, because it means your existing customers grow faster than your churning ones and your installed base compounds on its own. A 5% improvement in net retention compounds more than a 5% improvement in new logo acquisition, so I tell founders to fix retention before pouring money into acquisition.Should I scale before I have product-market fit?
No. Premature scaling is one of the most common reasons startups fail. Pouring money into sales and marketing before you have a product that retains customers is like pouring water into a leaky bucket. One company I worked with spent $500K on a sales team before confirming PMF and churned through three account executives, because the problem was product-market alignment, not sales execution.How does a founder's role have to change while scaling?
The leadership skills that take a company from $0 to $1M actively harm it from $1M to $10M if the founder can't adapt. Delegation becomes the job, systems thinking replaces firefighting, data replaces intuition, and culture stewardship becomes critical. A CEO who can't delegate becomes the ceiling on the company's growth.Can a small team build a large B2B SaaS company with AI?
Yes, and increasingly so. AI tools, no-code platforms, and remote work make it more feasible than ever to do more with a small, leveraged team. Across our portfolio we operate companies doing over $1M ARR with fewer than five full-time employees by trading payroll for an agentic stack, with humans focused on the work AI genuinely can't do, like vision, system design, product taste, and judgment.