I found a pricing page earlier where the most expensive tier looked rough until you switched the currency dropdown. Suddenly it looked affordable again. The same product, different currency, completely different psychology. It made me laugh, but it also sparked a question I can't let go of.
What if you could just tell an AI: "I want to sell these things"—and it generated your custom webshop?
It's a fun technical challenge. I've been playing with the idea of a "What do you want to sell?" website that feeds directly into an AI-powered shop generator. Custom implementation, all yours. But here's the honest part: I'm not sure there's a market for it, even if the tech works.
The math is simple. You can get a Shopify clone for less than 200 DKK a month. If you don't have anything yet, why wouldn't you start there? It's low friction, low risk, and it works. You're not going to bootstrap your e-commerce empire on a custom build.
But here's where it gets interesting: what if you did grow out of Shopify? What if you needed custom functionality that the standard platforms won't give you?
That's where an AI-built shop at 400 DKK a month becomes attractive. Custom implementation, fully yours. An agent developer who understands marketing, conversion funnels, payment integrations, inventory systems—all the things that actually matter. And as you scale and revenue starts flowing, you can upgrade the tier. Just like Shopify, but actually yours.
The maintenance stops being a burden because an agent can handle patches, updates, new features. That automation is baked in. You don't need to hire a developer every time you want to change how your product looks or how checkout works.
So the real question I'm sitting with: who actually wants a custom webshop?
Is it the person who just has five products and wants it to "work"? Probably not—they'll use Shopify. Is it the mid-market shop owner who's hit the ceiling of what standard platforms allow, but doesn't want to drop 50K on a custom build? Maybe. Is it the founder who wants full control over their tech stack because they know that's an unfair advantage in their market? Absolutely.
But is there enough of them to build a business around? That's the question keeping me up.
The technical side—agents building and maintaining your shop—that's solved. The business side—finding the market gap between "cheap but limited" and "expensive custom build"—that's the real challenge.
Maybe the answer is somewhere in the middle. Maybe it's not about building webshops at all. Maybe the insight is just that we're at this weird inflection point where an agent can handle things humans used to need to stay employed for.
Worth exploring though.
Part of the #100DaysToOffload documenting agentic development in 2026