Use this to compare candidate products side by side rather than evaluating one idea in isolation. Open the scorecard → (interactive scoring tool — in development; in the meantime, apply the manual scoring method below, and see How to Pick a Niche for the underlying logic).
Why score instead of just "deciding"
A single idea evaluated in isolation almost always looks appealing — you naturally focus on its strengths and rationalize its weaknesses once you're already interested. Scoring several candidates against the same fixed criteria at the same time forces an apples-to-apples comparison and surfaces which idea is actually strongest, not just which one you've spent the most time thinking about.
The manual scoring method
Score each candidate 1-5 on each dimension below, then compare totals. Weight the dimensions if some matter more to your situation than others (for example, weight margin more heavily if capital is your binding constraint).
| Dimension | What to score | Where to find the data |
|---|---|---|
| Demand | Search volume, review velocity on comparable listings, trend direction | Demand Research |
| Competitive strength of incumbents | Review count/rating of top listings, listing quality, seller sophistication | Competitive Research |
| Margin after realistic costs | Modeled net margin including fees, ads, and returns, not just landed cost | Product Profitability Calculator |
| Differentiation feasibility | Is there a specific, buildable improvement over what exists today? | Recurring complaints in competitor reviews |
| Compliance/operational complexity | Regulatory requirements, category gating, special handling (hazmat, cold chain, sizing/fit returns) | Category-specific marketplace policy pages |
| Sourcing feasibility | MOQ, lead time, supplier reliability for this specific product | Supplier quotes |
Setting a go/no-go threshold
Rather than picking whichever candidate scores highest by default, set a minimum bar in advance — for example, a candidate needs at least a "3" on every dimension, not just a strong average, since a single very weak dimension (e.g., margin) can sink an otherwise-strong idea regardless of how well it scores elsewhere. This catches ideas that look good on average but have one fatal flaw.
Worked example
Three candidate products score, out of 5 on each dimension: Product A scores 4/2/5/3/4/4 (strong on demand and margin, weak competitive picture); Product B scores 3/4/3/4/5/3 (moderate across the board, easy to source); Product C scores 5/5/1/2/3/2 (huge demand and differentiation angle, but the margin math doesn't work and sourcing is difficult). A naive "highest average" read might favor A or C, but applying a minimum-bar rule of "no dimension below 2" eliminates C outright on margin, leaving A and B as the real finalists — and forces you to dig further into A's competitive weakness before committing.
Mistakes to avoid
- Scoring only the idea you're already excited about, rather than a real shortlist of alternatives.
- Averaging scores without a minimum-bar rule, letting one fatal flaw hide inside a good-looking average.
- Scoring from memory/gut feel instead of pulling real numbers from demand, competitive, and margin research first.
FAQ
How many candidates should I score at once? Enough to force real comparison — three to eight is a practical range for most sellers; fewer than that and you're not really comparing, more than that becomes hard to research thoroughly.
Should every dimension be weighted equally? Not necessarily — weight the dimensions that matter most given your actual constraints (capital, storage space, risk tolerance) rather than treating all six as equally important by default.