Notes
Practical writing on Shopify and custom storefronts, vendor lock-in, SaaS architecture, AI-accelerated development, and marketing that does not rely on luck.
Most of this decision is settled by three questions about your catalogue, your checkout and your team — not by comparing feature lists. Here is the honest version, including when Shopify is obviously right.
Every agency says it. Almost none of them mean anything checkable by it. Here are the seven questions that turn the claim into a test — ask them before you sign, not after the relationship sours.
The fear is reasonable: a lot of AI-assisted code being shipped right now is bad. The difference is not the model, it is whether a human still owns the architecture and reads every line before it ships.
Most MVP advice says ship fast and fix it later, and it is mostly right. But four decisions are effectively permanent once you have paying customers. Get those right and cut corners everywhere else.
"Post consistently and it will grow" is a strategy with no control surface. Here is the version with one: the three numbers that decide whether paid acquisition works, and what to fix before spending anything.
Thirty minutes on a call and you'll have a fixed price and a timeline — or an honest no.