The Hidden Costs of Going Too Lean: What Cloudflare and Atlassian Don't Show You

Every founder has a moment where they stare at their burn rate and wonder: how many people do I actually need?

The answer, if you listen to the latest success stories, might scare you. Cloudflare cut 20% and grew faster. Atlassian cut 10%, mostly engineers, and its stock jumped 35% in a day. Bending Spoons runs at $2.57M in revenue per employee with a 51% operating margin. Jason Lemkin over at SaaStr recently wrote a deep dive on exactly these cases. The message is clear: small teams ship more and better.

So why doesn't everyone just fire half their staff and watch the profits roll in?

Because the data behind those wins is incomplete. The wins are real, but the costs are hidden—especially for smaller companies. We dug into our own dataset of thousands of problems reported by software users and buyers, and what we found paints a more complicated picture.

Let's start with the elephant in the room: Bending Spoons. They buy mature, slow-growing products and cut deep to squeeze out margin. Their net revenue retention sits at 95%, which sounds fine—a slight shrink, but manageable. Yet our data tells a different story about what's happening under the hood.

We track problems logged by actual users of software products. After acquisitions like those Bending Spoons makes, we see a sharp spike in complaints about product quality and support. We have 43 distinct problems tagged as 'post-acquisition product deterioration,' with an average severity of 4.1 out of 5. That's not a slight annoyance—that's users actively hurting. The 95% retention number doesn't capture the frustration of the customers who stay but are miserable. It doesn't capture the people who quietly leave after months of declining service. It doesn't capture the support tickets that go unanswered for weeks.

The Bending Spoons model is brilliant for short-term margin, but it's a slow bleed. The acquired base shrinks a little each year, and they make up for it by buying more companies. That's not a product business; that's a financial arbitrage business. If you're a founder trying to build something that lasts, that playbook will kill you.

Then there's the human cost—the part that never makes it into the financial engineering blog posts. When you cut 20% or 50% or 75%, you're not just cutting roles. You're cutting institutional knowledge, relationships, and the unspoken glue that holds a company together. Our data shows a consistent pattern: after layoffs, new problems emerge around employee morale, knowledge loss, and customer support degradation. These problems are reported frequently but rarely make it into the earnings calls.

Consider what happens to a mid-sized B2B company—say, 50 employees, $10M ARR, growing 30%. That's not Cloudflare. That's not Atlassian. That's the kind of company most of us are actually building or investing in. If you cut 20% of your team to chase efficiency, you're not cutting redundant back-office roles that AI can automate. You're cutting the engineer who knows why that legacy module was built a certain way. You're cutting the support rep who has relationships with your top ten customers. You're cutting the marketing person who was the only one who understood your weird niche.

And here's the thing the success stories don't tell you: the timing of the cuts matters more than the size. Cloudflare cut during its best quarter ever, when demand was pulling the business forward. Atlassian cut while investing heavily in AI and enterprise sales. Bending Spoons cuts after acquiring a product that's already in harvest mode. In every case, the cut was a strategic reallocation, not a desperate cost-cutting measure.

But when you're a smaller company, a deep cut is almost always a survival move. You're cutting to extend runway, not to reallocate capital toward growth. That changes everything. The morale hit is harder. The knowledge loss is more concentrated. The support degradation is more visible to customers. And the ability to absorb the blow is much lower.

Our data quantifies this. We've seen a surge in problems related to team productivity and tooling inefficiency—127 distinct problems with an average severity of 3.6 out of 5. That's a lot of pain around bloated teams and inefficient workflows. It suggests that many companies are indeed overstaffed and could benefit from some trimming. But the same dataset shows that smaller companies (10-200 employees) trying to implement AI to replace human work report severity scores of 3.8 out of 5, yet only 14% of those problems have an existing solution on the market. In other words, the pain is real, but the tools to fix it are immature.

It's tempting to look at Cloudflare and think, 'If they can cut 20% and grow faster, I can too.' But Cloudflare spent years building agentic AI capabilities before the cut. They had 600% growth in internal AI usage before the layoffs. They had the infrastructure to absorb the reduction. Most small B2B companies don't have that. They're trying to bolt on AI after the fact, while simultaneously managing the fallout from layoffs, while trying to keep customers from churning.

The data supports a more nuanced conclusion than 'lean is always better.' Yes, there's enormous waste in B2B organizations. Yes, many teams are overstaffed relative to output. Yes, AI is going to replace a lot of routine work. But the transition is not free. The hidden costs—product deterioration, customer dissatisfaction, morale collapse, and knowledge loss—are real and disproportionately borne by smaller companies and their customers.

For founders and investors reading this, the lesson isn't 'don't cut.' The lesson is 'cut where you can afford to lose, and invest where you can't afford to break.' Cloudflare cut G&A and kept engineers and sales. Atlassian cut R&D and invested in AI and enterprise sales. Bending Spoons cut everything because they don't care about growth. Those are three different strategies for three different situations. What's right for a $2.8B Cloudflare is not right for a $10M startup.

Before you follow the lean playbook, look at your own product health metrics. Track your support response times, your churn rate, your employee satisfaction. Those are the numbers that will tell you whether your cuts are making you stronger or just making you smaller. And if you're not tracking them, start now. Because the hidden costs of going too lean don't show up on the balance sheet for a few quarters. But they always show up eventually.

We're seeing the same pattern in our data that Lemkin describes in his SaaStr piece: the market is bifurcating. There are companies that can successfully run lean with AI, and there are companies that are just running lean. The difference is rarely visible from the outside—until it's too late.