The Niche Test: How to Pick Work That Pays Double

“Niche down” is the most repeated and least useful advice in freelancing. Nobody tells you how to choose, so most people either stay a generalist or pick a niche based on what sounds impressive. Here’s a filter that actually works.

The three questions

A good niche has to pass all three. Two out of three isn’t enough.

1. Do they have money and a budget for this?

Passion markets are a trap. You want clients for whom your work is tied to revenue, not a hobby. Ask: when this problem is solved, does someone make or save money? If yes, there’s a budget.

2. Can you reach them in one place?

A niche you can’t find is a niche you can’t sell to. If your ideal clients gather somewhere — a conference, a subreddit, a trade association, a software ecosystem — you can market once and reach many.

3. Is the problem expensive enough to specialize over?

If the problem is trivial, clients won’t pay a premium for a specialist. The sweet spot is a problem that’s painful, recurring, and currently handled badly by generalists.

Test before you commit

You don’t have to bet the business on a guess. Run a 30-day test:

• Rewrite your headline for that one niche.
• Reach out to ten people who fit it.
• Track reply rate and how fast money comes up in the conversation.

If the conversations get serious faster than your generalist pitch ever did, you’ve found something. If they don’t, you’ve spent 30 days, not a year.