Adapt · Essay

Stop trying to turn the mothership.

Greg Moran · November 2026 · 5 min read

Legacy SaaS companies have plenty of talent and plenty of capital. They are still losing to five-person teams. The problem is not the technology. It is the organization wrapped around it.

The swamp

Most legacy SaaS companies are trying to build a new business model inside an org designed for the old one. Seat-based pricing. Roadmaps owned by existing customers. Codebases nobody wants to touch. Decisions that take a quarter.

Meanwhile, AI-native startups ship in weeks. Every month spent transforming the core is a month further behind, and the gap compounds. Eventually, and quickly, it becomes insurmountable.

Here is the part most boards have not caught up to. The biggest liability in legacy SaaS is not technical debt anymore. It is organizational debt. Refactoring code takes a quarter. Refactoring an org takes years, and the market is not going to wait for you to finish.

Build new ships

So stop trying to turn the mothership. Build new ships.

Run it like an internal venture studio. Teams of two to five people. Their own codebase and their own P&L. One customer problem each. Ship it, sell it, watch the numbers.

No signal? Kill it cleanly. Real signal? Lean in hard.

The parent company brings capital, distribution, domain expertise, and a customer base to sell into. That is a genuine edge no startup has. What the parent does not bring is its operating model. The moment the new team inherits the approval chains, the pricing committee, and the quarterly roadmap review, you have rebuilt the mothership in miniature and you will get the mothership’s speed.

Structure matters less than autonomy

People get stuck on the legal question. Studio, separate business unit, eventual spinout? The structure matters less than people think. Autonomy is what drives it. The test is simple: can the small team change its price, its product, and its priorities this week without asking anyone in the core business?

If the answer is no, you do not have a new ship. You have a committee with a code repository.

What I would do Monday

1. Pick one customer problem your core product is too slow to solve. Not a feature gap. A problem a customer would pay for on its own.
2. Pull two or three people off the roadmap. The ones who keep asking why things take so long. Give them their own stack and their own number.
3. Set a 90-day bar: something shipped, something sold. Revenue is the only review that counts.
4. Decide the kill criteria in advance. If the signal is not there at 90 days, kill it cleanly and start the next one. The cost of the experiment is the point. It has never been lower.

The honest tradeoff

This approach creates friction. The core team will resent the new team’s freedom. Some experiments will die in public. And the new ships will eventually cannibalize core revenue, which is uncomfortable right up until you remember that someone else’s startup was going to do it anyway.

If you run legacy SaaS, this may be your only way out. The companies that make it will not be the ones that transformed the core. They will be the ones that launched ships from its deck while the core was still paying the bills.