A lot of sellers pick a niche, spend a weekend making 20-30 designs, upload everything, and then wait. That's backwards. You can find out whether a niche has real demand before you've made a single design, and it takes maybe an hour.
Start with Etsy's own search bar. Type in a broad term related to your idea — say, "cottagecore mushroom" — and look at what autocompletes. Those suggestions aren't random; they're built from actual search volume. If Etsy is suggesting "cottagecore mushroom shirt," "cottagecore mushroom sticker," and "cottagecore mushroom tote," that's a decent signal people are searching those exact phrases.
From there, actually search the term and sort by "Best Selling" instead of "Relevancy." This is the part most people skip. You want to see what's already working, not just what exists. If the top 20 results all have hundreds or thousands of sales, there's proven demand — your job is differentiation, not invention. If the top results have single-digit sales after being listed for a year, that's a sign the niche is smaller than it looks, or the existing listings are all just bad at marketing (less common, but worth checking).
Google Trends is worth five more minutes here. Pop in your niche term and look at the 12-month trend line, not just the current number. A term that's flat or growing is safer to build around than one that spiked and is clearly sliding down — you don't want to spend two weeks building a collection around something that peaked three months ago.
The last check, and honestly the one people skip most often: look at who's actually buying in that space. Scroll through reviews on the best-selling listings you found. Are people buying these as gifts? For themselves? Mentioning a specific occasion? That tells you not just whether the niche works, but how to market it once you're in it — what to put in your titles, what season to launch in, what your product photos should look like.
None of this takes more than an hour, and it's the difference between building 20 designs into a proven direction versus 20 designs into a guess. If a niche fails all four checks, that's not necessarily a dead end — it might just mean you need a sharper angle within it rather than abandoning it entirely.