Every “50 profitable KDP niches” list on the internet is the same list. Gratitude journals, wedding planners, sudoku — copied from last year’s version of the same article, with no numbers behind any of it.
This one is different in a checkable way. The 16 niches below come from KDPie’s live research canon — several hundred niches scanned with real Amazon data. For each one you get the measured demand signal, the actual competitor picture, and the verdict our locked scoring formula assigned: pilot (enter it) or explore (real signal, verify the angle). Every row was re-scored under the current formula on August 12, 2026, the day before publishing. One row changed its verdict during that re-score — upward, as it happens — which tells you something about lists that never re-check anything.
Full disclosure: KDPie is our product, and yes, this article is also a demonstration of what it does. The numbers are real either way.
How to read the table
- Demand — measured Amazon search volume for the niche’s head term (searches on Amazon, not Google), plus how many of the visible competitors are actually selling (Best Sellers Rank under 50,000 — see what BSR means). Some strong niches have no measurable head-term volume; demand then shows in the selling competitors.
- Competition — how many of the visible competitors have under 30 reviews, i.e. are beatable by a well-made newcomer.
- Verdict + score — from a fixed, deterministic formula (same data in, same verdict out). Not our opinion, and not adjustable to make a niche look better.
And one honest note: these are not our 3 highest-scoring niches. The very top of the canon stays in the product — the three strongest are health journals scoring 82–84 that paying users can find in one scan. What’s below is still strong: eight pilots and eight explores, chosen to show the patterns worth copying.
The 16 niches
| Niche | Market | Verdict | Score | The evidence in one line |
|---|---|---|---|---|
| Kita educator planner (erzieher planer für die kita) | DE | pilot | 79.5 | 38 competitors selling well; 32 of 40 beatable (<30 reviews) |
| Endometriosis pain journal (endometriose schmerztagebuch) | DE | pilot | 79.0 | head term ~748 searches/mo; 20 of 27 competitors beatable |
| Teacher-trainee planner (referendariat planer) | DE | pilot | 78.3 | 8 selling well; no competitor above 100 reviews |
| Large-print sudoku for seniors (großdruck sudoku) | DE | pilot | 76.0 | 10 selling well; only 1 competitor above 100 reviews |
| Large-print word search for seniors (wortsuchrätsel großdruck) | DE | pilot | 75.6 | ~635 searches/mo; 25 of 36 beatable |
| Nonogram puzzle book (nonogramme rätselbuch) | DE | pilot | 74.9 | 17 selling well; no 500+ review incumbent |
| Str8ts puzzle book | DE | pilot | 74.8 | 21 selling well on a ~112/mo head term |
| Futoshiki puzzle book | DE | pilot | 74.6 | 28 of 31 competitors beatable; nobody above 100 reviews |
| Model-railway build log (modellbahn bautagebuch) | DE | pilot | 71.5 | head term ~1,533 searches/mo; only 1 competitor selling well |
| Puppy’s first-year journal (welpentagebuch) | DE | explore | 69.4 | tiny measured demand (~37/mo) but 29 of 31 beatable |
| Yoga teacher class planner | DE | explore | 68.2 | 7 selling well; 22 of 30 beatable |
| Blank sheet-music notebook (notenheft blanko) | DE | explore | 65.8 | ~299 searches/mo but a crowded field — angle needed |
| Celiac / gluten-free food diary | DE | explore | 62.4 | head term ~1,084 searches/mo; 31 of 44 beatable |
| RV camping logbook | DE | explore | 60.2 | 13 selling well, but 1 incumbent with 500+ reviews |
| Dog walker daily log (for pet sitters) | US | explore | 57.9 | 27 of 29 beatable; demand not measurable — verify first |
| Cleaning business client schedule book | US | explore | 55.3 | same shape: wide-open shelf, unproven demand |
The patterns — the part you can actually reuse
The list matters less than what repeats inside it. Four patterns carry almost every niche above:
1. Audience × format beats format alone. “Sudoku” is dead — a shelf owned by thousand-review incumbents. “Large-print sudoku for seniors” scores 76 as a pilot, because a specific buyer with a specific need re-opens a closed market. The same move works twice more in the table (word search and logic puzzles, both large-print). One warning from our own data: the specificity has to be real. The formula discounts broad head terms precisely because “journal” demand says nothing about “your journal” demand.
2. The second-tier puzzle move. Sudoku is saturated; str8ts, futoshiki, nonogramme are not — smaller audiences, but shelves where 90% of competitors have under 30 reviews and nobody has built a brand. Puzzle buyers who finish one book buy the next one. Three pilots in the table come from this single move.
3. Profession log books. Kita educators, teacher trainees, yoga teachers, dog walkers, cleaning businesses: professions with a documentation habit and no dominant supplier. These niches look tiny from the outside — that is exactly why the shelves are open. The two US rows show the pattern’s risk end: wide-open shelf, but demand you must verify before building.
4. Condition journals. The endometriosis pain journal — the row that flipped explore→pilot on re-score — and the celiac food diary both ride a real head term (748 and 1,084 searches/mo) into an under-served sub-audience. People managing a condition track things, and generic trackers serve them badly. Handle this category with respect for the reader and it converts.
Why so much Amazon.de in an English article
Because that is where our canon is currently deepest, and because it makes the underlying point better than a US list would: the non-US Amazon stores are the under-tooled part of KDP. Nearly every research tool is US-first; German, French and Spanish shelves get a fraction of the competition per unit of demand. You publish there through the same KDP account with zero extra friction. If you only ever research Amazon.com, rows like “model-railway build log: 1,533 searches/mo, one competitor selling well” are what you’re not seeing.
The patterns transfer both ways, and translating a niche phrase is not translating a niche — demand must be re-measured per market. Which is a thing tools should do for you, not a thing to assume.
Before you build any of these
A published list ages from the day it goes up — including this one. Treat the table as evidence of where to look, then verify the specific shelf the week you commit: measure the phrase (how), read the top 10’s ranks (how, or use the free BSR calculator), and check the money at the going price with the royalty calculator. And if you’re reading this in the second half of the year, check the Q4 timing math first — some launches no longer have time to rank before October.
Or let the machine do exactly that: every verdict in this table came out of a KDPie scan anyone can run — the $7 trial covers enough scans to check your own shortlist, on whichever Amazon market you publish in.