Thoughts — Dominic Emanuel Horn

I'm calling it: a lot of SaaS products are existentially threatened by recent advances in AI coding. Not in some abstract, five-years-from-now way. Right now.

Enterprises are hiring and enabling teams of vibecoders to create internal replacements to things like Trello, Typeform, Miro, Canva, Notion, and others. Why pay tens of thousands of dollars a month for per-seat software if you can replicate the functionality you actually need at a fraction of the cost?

Will it be as good as the dedicated software? Probably not. Not yet. But it will be good enough, tailored to your exact business needs, keep all data internal, and save boatloads of money.

That last part is the quiet killer. "Good enough" has destroyed entire industries before.

The 80% you actually use

Here's the thing nobody talks about: most companies use maybe 20% of what a SaaS product offers. You're paying for Notion's databases, its AI features, its API, its publishing capabilities, its team permissions matrix — and your team uses it as a glorified wiki with a few linked pages.

You're paying Miro enterprise rates so people can put sticky notes on a canvas once a quarter during a strategy offsite.

A vibecoder with good product instincts can look at how your team actually uses a tool and build exactly that. Nothing more, nothing less. No feature bloat. No "we're sunsetting the feature you depend on." No surprise pricing changes when they decide to chase upmarket.

The result is something lean, fast, and shaped like your business — not shaped like a VC-backed company's attempt to be everything to everyone.

The real math

Let's make this concrete. Say you're a mid-size company, 200 people, running a fairly typical SaaS stack:

Project management tool: €15/user/month
Form and survey tool: €300/month
Whiteboarding tool: €10/user/month
Design tool: €12/user/month
Knowledge base: €10/user/month

Even if only half your team is on each tool, you're burning €30,000–50,000 a month on software subscriptions. That's €360K–600K a year.

Now imagine you put one or two vibecoders on retainer. Maybe €10K–15K a month total. They spend their time doing what a tiny internal product team would do: researching what your people actually need, building tailored tools, iterating based on real feedback. Within six months you've replaced three or four subscriptions with internal tools that do exactly what you need and nothing you don't.

The savings pay for the vibecoders several times over. And you keep building from there.

Internal micro-startups

I see a real opportunity forming here that most companies haven't recognized yet: the vibecoder on retainer.

Someone who operates like a one-person internal startup. They identify problems, research existing solutions, prototype replacements, and ship working software. Continuously.

Think of it as a product person who can also build. Someone with high product intuition who happens to have AI-accelerated development capabilities. That combination is rare right now, which means it's valuable.

The best part? These internal tools compound. The form builder your vibecoder creates in month one becomes a building block for the customer feedback system they build in month three. The internal dashboard they ship in month two becomes the foundation for a client-facing portal in month six. Unlike SaaS subscriptions, which are pure cost, internal tools are assets that grow.

The data argument

Every time you put company data into a third-party SaaS product, you're making a trust decision. Most companies make that decision on autopilot because "everyone uses Notion" or "Slack is industry standard."

But here's what's shifting: enterprises are getting increasingly nervous about where their data lives, who trains models on it, and what happens when a SaaS provider gets acquired or pivots. The compliance overhead alone — vendor assessments, security reviews — costs time and money that rarely shows up in the SaaS budget line.

Internal tools sidestep all of that. Your data stays on your infrastructure. Your compliance team sleeps better. Your legal department has one fewer vendor contract to review. For regulated industries — finance, healthcare, legal — this isn't a nice-to-have. It's becoming the default argument.

What SaaS companies should be worried about

If I were running a SaaS company right now, I'd be asking myself one hard question: what's my moat?

Because it's no longer "we built software that does X." A vibecoder can build software that does X in a weekend. The question is what else you bring to the table.

Network effects are a moat. Figma is hard to replace not because the design tool is irreplaceable, but because the collaboration layer and community ecosystem took years to build. Slack is hard to replace because the integrations are the product.

Deep domain expertise is a moat. A vertical SaaS for, say, dental practice management has years of regulatory knowledge and workflow optimization baked in that a vibecoder won't replicate quickly.

Data and intelligence are a moat. If your product gets meaningfully better the more people use it, that's defensible.

But if your product is essentially a CRUD app with a nice UI and some collaboration features? That's the blast radius. And it's bigger than most founders want to admit.

The surprising part: this helps SaaS companies too

Here's the counterintuitive angle. The smart SaaS companies will use this same dynamic to their advantage. Instead of fighting the vibecoders, they'll become platforms.

Offer APIs and SDKs that make it easy for internal tools to integrate with your product. Become the backbone, not the UI. The companies that survive won't be the ones trying to lock users into their interface. They'll be the ones that make themselves indispensable as infrastructure.

Stripe doesn't worry about vibecoders. Neither does Twilio. They're too deep in the stack to replace. The SaaS companies at risk are the ones sitting at the top of the stack, where the UI is the product.

Who should be paying attention

If you're a business spending €20K+ a month on SaaS and you haven't explored what an AI-accelerated builder could replace — you're leaving money on the table. Probably a lot of it.

If you're building SaaS and your differentiation is "we have a nice UI for [common workflow]" — start thinking about what happens when that stops being enough.

And if you're someone with product sense who's been learning to build with AI tools — the market for what you can do is about to expand dramatically. You're not just a developer or a designer anymore. You're an internal product studio in one person.

Your business won't subscribe to some pay-per-user SaaS anymore. Instead, you'll have dedicated, tailor-made software, built for your business.

If you're a business that wants to replace some expensive niche tool you're using? Set up a call. Maybe we can make it happen.