Where did the SaaS moat go?

I’ve just built a decent software-as-a-service (SaaS) application in a few weeks. A productivity tool with multiple screens, workflows, user management. The kind of thing that would have taken a small team months not long ago. I did it largely with Claude Code.

That should worry a lot of people.

The personal experiment

Boudicaa started as a personal project. A Getting Things Done app rebuilt from an old PHP codebase using AI coding tools. But somewhere during the process I realised I was looking at something that could credibly be offered as a SaaS product. Not because I’m exceptional. Because the tools have made this level of output accessible to any experienced developer working alone.

The moat that protected software businesses used to be the sheer expense and difficulty of building the product. That moat is draining fast.

The market agrees

I’m not the only one who’s noticed. The Economist recently reported that listed software firms have lost around a fifth of their value this year. Investors expect that AI coding tools will allow customers and startups to undercut incumbent software businesses. Five years ago, the largest software firms in the S&P 500 were valued at around thirteen times their annual revenue. That figure has dropped to roughly eight. Across all listed American software firms, the multiple has fallen from eight to three.

Private equity is particularly exposed. During the 2010s, buy-out funds poured roughly one dollar in three into technology firms. Those deals were struck at high valuations and funded largely by debt. Analysts estimate that more than $500bn in borrowing tied to software firms sits in American credit markets. Some lending funds with heavy software exposure are already seeing investors head for the exits.

Meanwhile on the Jagged Frontier

I built Boudicaa for myself. It works well as a personal productivity tool. I could market it as a service. But launching a SaaS product requires infrastructure, support, marketing, compliance. Costs that might once have been covered by venture capital. That funding looks a lot less likely now. Investors can see that the next developer with Claude Code can build something similar in weeks. With the barrier to entry collapsing, why would a VC back a product with no moat?

Design still matters when code is cheap. But cheap code also means cheap competition. Every SaaS product that relies on its codebase as its competitive advantage is now vulnerable.

What to do?

That’s the billion-dollar question which a lot of people are working on right now. The code isn’t the moat anymore. Maybe the value has moved to domain expertise, data, relationships with users. Maybe the software becomes a commodity and what sits around it becomes the differentiator. I don’t know yet.

The private equity firms that bet billions on software margins are learning this from the top down. I’m learning it from the bottom up, one app at a time.

Where do you think the moat moves to when anyone can build the software? I’d love to hear from founders and developers rethinking their competitive advantage.