Merch by Amazon niche research: a workflow you can repeat every week
Good Merch niches are not found by scrolling best-seller lists for hours. They come from a short routine: start from a buyer, prove that buyer is spending, see how crowded and how fast the niche is moving, and clear the words before you design. Here is that routine, step by step.
Updated 5 min readNicheKite
What a niche means on Merch
On Merch by Amazon a niche is not a product category. It is a group of buyers who share an identity and will wear it: night-shift nurses, retired firefighters, people who own a corgi, kindergarten teachers in their first year. Shirts sell when the buyer reads the design and thinks “that's me” or “that's my dad”.
That definition matters for research. You are not asking “do shirts sell?” You are asking three narrower questions: is this group buying shirts about itself on Amazon right now, how many designs already serve it, and is interest growing, stable or seasonal.
Step 1. Start from audiences and dates, not from designs
Write a list before you open any tool. Two sources give you most ideas:
- Evergreen audiences. Jobs (nurse, welder, accountant), hobbies (fishing, crochet, disc golf), family roles (new grandpa, bonus mom), pets and breeds, places, life events (retirement, graduation). These sell all year with peaks around gift dates.
- Dates. Holidays and observances create short, predictable demand: Mother's Day, Father's Day, Halloween, Christmas, but also smaller ones like Nurses Week or National Cat Day. Pick dates six to ten weeks ahead so a new design has time to be approved, indexed and earn its first sales.
Cross the two lists: “Father's Day × fishing”, “Halloween × teachers”, “Nurses Week × night shift”. Each pair is a niche candidate. The NicheKite calendar lists 4,488 US, UK and German dates with a short note on who buys for each, which is a fast way to fill the date side.
Step 2. Prove demand with Best Sellers Rank
Amazon does not publish unit sales, so Merch sellers use Best Sellers Rank (BSR) as the proof of demand. A design only gets a rank after it sells, and a lower number means more recent sales. Search the niche and keep only designs that have a rank.
Then read the first 20 to 40 results like a buyer would, and note four things:
| What to read | Why it matters | Healthy sign |
|---|---|---|
| Rank of the top designs | Shows whether the niche sells regularly or only now and then | Several designs with a strong rank, not just one |
| Listing dates | Tells you whether new designs can still break in | Some designs in the top results were listed in the last months |
| Prices | Shows what buyers accept and where you can position | Designs selling at or above the default price |
| Brand | Licensed brand stores rank on brand power, not on the niche | Most of the top results are independent sellers |
If you want to go deeper into what different rank levels mean, read how to read BSR on Merch by Amazon.
Step 3. Measure competition
Demand alone is not enough. A phrase used by thousands of designs can sell well and still be a poor bet for a new listing. Two numbers help:
- How many designs match the niche. The result count of a ranked-only search, compared with the count without the BSR filter, tells you how many listings exist and how many of them sell.
- How many designs use the phrase, and whether that number is growing. Keyword analytics counts the designs using each phrase and plots the change. A broad word like “teacher” returns longer phrases (substitute teacher, math teacher, sped teacher) that are smaller sub-niches.
The sweet spot is a phrase where ranked designs exist (demand) while the number of designs is flat or falling (competition is not piling in).
Step 4. Read the direction: rising, stable or fading
A best-seller list is a picture of what already won. To see where a niche is heading, look at rank change over time. In NicheKite, sort any search by Fastest rising to see which designs gained the most rank in the last 30 days, or open Trend Hunting for the whole marketplace.
Read the pattern rather than any single design. When several designs on the same theme rise together, buyers are discovering the theme. When only one rises, it is usually that design's moment, not the niche's.
Step 5. Check last year for seasonal niches
Seasonal niches repeat, and last year is the best forecast you have. Before you design for a date, open Merch Archives at the same window last year (one to three weeks before the date), filter to a strong rank band and sort by trend. You get the themes that actually sold, not the ones people talked about.
Write down themes, not artwork: the phrases, the audiences, the tone (funny, proud, sentimental) and the styles that repeat. Copying a design is both against Amazon's policy and a poor strategy; the theme is what transfers.
Step 6. Clear the words before you design
Many rejected or removed Merch listings fail on words, not on art. Phrases that look generic can be registered trademarks for clothing (USPTO class 25). Check your planned title, both bullets, your brand name and any text in the design before you spend time on artwork.
The trademark checker highlights words that match live clothing trademarks in red and words that belong to broader marks in yellow. The full routine, including what the checker cannot see, is in how to check trademarks for Merch by Amazon.
A simple scorecard for each niche
Score every candidate the same way so you can compare them on a Friday afternoon and decide what to design next week.
| Question | Evidence | Score 0–2 |
|---|---|---|
| Is it selling? | Several independent designs with a strong rank | 0 none · 1 a few · 2 many |
| Can new designs get in? | Recently listed designs in the top results | 0 none · 1 one or two · 2 several |
| Is it crowded? | Design count for the phrase and its trend | 0 growing fast · 1 stable · 2 flat or falling |
| Where is it heading? | Fastest rising inside the niche, last year's archive | 0 fading · 1 stable · 2 rising or seasonal peak ahead |
| Are the words clear? | Trademark check of title, bullets and design text | 0 red match · 2 clear |
Common mistakes
- Chasing licensed brands. Official stores rank on brand demand. Filter them out and study independent designs.
- Designing for the date, too late. A Halloween design uploaded in late October rarely ranks in time.
- Treating one viral design as a trend. Wait for the pattern across several sellers.
- Researching the art and ignoring the words. The title and bullets are what get listings found and what get them removed.
- Never revisiting. Re-run saved searches weekly. Niches you passed on can open up when the competition fades.
Questions
How many niches should I research before designing?
Enough to fill your next batch of uploads with candidates that scored well. Many sellers research in batches of 10 to 20 niches and design for the top few.
Is a niche with many designs always bad?
No. A large niche with steady demand can still work if you target a narrower sub-niche inside it, which keyword phrases help you find.
Do I need paid tools for niche research?
You can do every step by hand on Amazon and the USPTO website. Tools save the time of collecting rank history, counting designs and checking trademarks one phrase at a time.