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.