What a site starter actually buys you

Every few months somebody asks me whether they should buy a site starter or build from nothing. The honest answer is that it depends on which part of the work you actually enjoy. What follows is how I decide, what a starter is good at, where it quietly costs you, and how to tell a careful one from a careless one before you pay.

What a starter is

A starter is a complete, working site handed to you as source code: pages, navigation, typography, a content pipeline and a build that already produces something deployable. You are not buying pixels. You are buying the two hundred small decisions somebody already made and tested — where the reading column ends, how the mobile menu traps focus, what the heading scale does at 1440 pixels and at 375.

That framing matters, because it tells you what you are still on the hook for. A starter does not know your writing, your projects or your audience. It gives you the scaffolding and expects you to bring the substance.

Getting your money’s worth

The failure mode I see most often is treating a starter as a finished product. People change the name, publish, and then never touch it again — so the site reads like somebody else’s, because it is. The starters that pay off are the ones people edit within the first week.

Start with the content modules rather than the components. Almost every good starter keeps its copy in a handful of plain data files, which means you can replace every word on the site before you learn the component tree.

// src/data/site.js — the first file worth editing
export const site = {
  name: 'Your Name',
  description: 'One sentence a stranger would remember.',
  socials: [{ label: 'Code', url: 'https://example.com' }],
};

Once the words are yours, the layout starts to look wrong in useful ways. That is the signal to open components — not before.

Selling one of your own

If you have shipped a few sites, packaging one is a reasonable side income and a very good forcing function. Nothing exposes sloppy work faster than knowing a stranger will run install and build on a machine you have never seen.

The parts buyers complain about are never the visuals. They are missing documentation, hard-coded personal data buried three components deep, a dependency that needs a native compiler, and animations with no reduced-motion path. Fix those four things and you are ahead of most of the market.

What a careful starter looks like

Read the source before you read the screenshots. I look for four things: content separated from markup, internal links that survive being served from a subpath, interactive parts that work with a keyboard, and a build with as few dependencies as the job allows. A starter with two dependencies and clear data files will outlive one with forty and a clever animation on every scroll.

Also check the boring surface: does the mobile menu close on Escape, does the contact form fail gracefully, is there an obvious place to put your own artwork. Those are the details you will meet on day two.

Conclusion

A starter buys you decisions, not identity. It is worth the money when the part you want to spend time on is the writing, the projects or the launch — and a poor trade when what you actually wanted was to learn how the layout works. Buy one, gut the content files in the first sitting, and keep the parts that were better than what you would have written yourself.