"Follow your passion" is incomplete advice for picking a niche — plenty of passion projects fail on unit economics. A better framework scores each candidate niche on three independent dimensions.
Dimension 1: Demand
Does a real, searchable audience already want this? Evidence: marketplace search volume, existing competitor sales volume (visible via review counts as a rough proxy), and general web search interest. A niche with zero visible demand isn't necessarily a "blue ocean" — it's more often a niche with no market.
Look beyond a single data point. Cross-reference marketplace search suggestions (the autocomplete terms a marketplace's own search bar surfaces), general web search interest over time, and social platform activity (is the category showing up in organic content, not just ads). Directional agreement across two or three of these sources is a much stronger demand signal than any one alone.
Dimension 2: Competition
How crowded is it, and — more importantly — is the competition beatable? A crowded market with mediocre existing products (visible in review complaints) is more attractive than a smaller market fully dominated by one strong, well-reviewed incumbent brand.
Ask specifically: is the top listing's success built on genuine product quality (hard to beat), or on being first-to-market with an otherwise unremarkable product (beatable with a better offering)? Read the 2-3 star reviews on the top 3-5 listings — they're usually more informative than 5-star or 1-star reviews for finding a genuine gap, since they tend to describe specific, fixable shortcomings rather than either pure praise or an isolated bad experience.
Dimension 3: Margin potential
Can you get a landed cost low enough, at a sourcing volume you can actually afford, to hit a healthy contribution margin after marketplace fees, shipping, and a realistic ad spend? Run real numbers through the Unit Economics Calculator before falling in love with a niche.
Watch for categories that look attractive on demand and competition but are structurally thin on margin — commonly, products with a low average selling price relative to their bulk/weight (expensive to ship relative to what they sell for), or categories where the "winning" price point sits uncomfortably close to your landed cost once realistic advertising spend is included.
A simple scoring approach
For each candidate niche, score 1-5 on demand, competition (5 = weak/beatable competition), and margin potential, then multiply the three scores rather than adding them — a niche that's a 5 on demand and competition but a 1 on margin (i.e., structurally unprofitable) should score low overall, not just "a bit lower." Multiplication punishes a fatal flaw in any single dimension the way reality actually does.
Worked example
| Niche | Demand (1-5) | Competition (1-5, 5=weak) | Margin (1-5) | Score (multiplied) |
|---|---|---|---|---|
| A | 4 | 3 | 4 | 48 |
| B | 5 | 5 | 1 | 25 |
| C | 3 | 4 | 4 | 48 |
| D | 2 | 2 | 2 | 8 |
Notice Niche B looks the most exciting on demand and competition alone (a 5 and a 5), but a bottom-tier margin score drags its multiplied total down to tie with a much less flashy-looking Niche C. This is exactly the trap the multiplication approach is designed to catch — an additive score would have ranked B far higher than it deserves.
A note on passion and expertise
Passion and category expertise are real advantages — they help with content, customer empathy, and staying motivated through the hard early months — but they're a tiebreaker between niches that already score well on demand, competition, and margin, not a substitute for those three. If two niches score similarly on the framework, choose the one where your own knowledge gives you a real head start on differentiation or content.
When the framework says "none of these work"
Sometimes every candidate on your shortlist scores poorly, usually on margin. That's a legitimate outcome, not a framework failure — it means the shortlist needs to expand. Go back to broader category browsing, look at adjacent niches to your original ideas, and re-run the scoring rather than lowering your bar for what counts as a passing score.
Mistakes to avoid
Beyond the passion trap already covered, watch for anchoring too early on the first niche you research — score several candidates side by side so you have a real comparison rather than judging one idea against an abstract, unspecified standard.