How to Evaluate a Startup Idea: A Practical Scoring Framework

Every founder has more ideas than time. The hard part is not generating them but deciding which one deserves the next six months. Gut feel is biased toward whatever you thought of most recently or find most fun to build.

A simple weighted scorecard fixes most of that. It will not predict success, but it forces you to look at the same factors for every idea, exposes weak spots early, and makes comparisons honest. Here is a framework you can use in under an hour per idea.

The seven criteria

Score each idea on seven dimensions. The weights reflect a common heuristic for small, bootstrapped or early-stage teams: pain and distribution matter more than market size, because a small team with a desperate, reachable audience usually beats a small team chasing a huge but vague market.

Feel free to change the weights. A venture-backed team might weight market size higher; a solo founder with a day job might weight build effort higher. What matters is that you pick the weights before scoring, and use the same ones for every idea.

How to score each criterion from 1 to 5

Vague scales produce vague scores. Anchor each number to something observable.

Pain intensity

• 1: Mild annoyance; people shrug when asked
• 3: People complain and use workarounds, but rarely pay
• 5: People spend money or hours every week on it and ask for alternatives unprompted

Reachability and distribution

• 1: No clear channel; you would need broad paid ads
• 3: Communities exist but are hard to reach or wary of promotion
• 5: You know exactly where they gather and can reach them this week

Monetization

• 1: Users expect it free; no obvious buyer
• 3: Some willingness to pay, low price point
• 5: Clear budget owner, existing spend you can replace

Market size and timing

• 1: Tiny niche with no growth, or no reason it works now rather than five years ago
• 3: Reasonable niche, stable demand
• 5: Growing segment with a clear recent trigger such as a new platform, API or regulation

Competition and differentiation

• 1: Strong incumbents, and you cannot name a reason to switch
• 3: Competitors exist; you have a plausible angle
• 5: Competitors are weak, outdated or ignoring a segment you can own

Founder fit

• 1: No domain knowledge, no network, low interest
• 3: Some relevant experience or access
• 5: You have lived the problem and know people who have it

Build effort

• 1: Many months before anything useful ships
• 3: A few weeks for a basic version
• 5: A usable version in days, or no code needed at first

Multiply each score by its weight and add them up. The result is a number between 1 and 5.

Worked example (fictional idea)

To make this concrete, here is a made-up idea scored end to end. It is an illustration, not a recommendation.

Example idea: a tool that reads a dental clinic's appointment calendar and automatically texts patients to fill last-minute cancellations from a waitlist.

A 3.2 is a "maybe." The scorecard also tells you why: strong pain and monetization, weak competition and founder fit. The next step is not building; it is checking whether existing suites really cover this, and finding a few clinic owners to talk to. That is exactly what a 48-hour validation plan is for.

Red flags that override the score

Some problems are serious enough to downgrade an idea regardless of its total:

• You cannot name a single real person with the problem. The score is fiction until you can.
• "Everyone" is the customer. Ideas without a specific first audience have no distribution plan.
• The buyer and the user are different, and you only talked to the user. Teachers may love it; the school district pays.
• The core depends on a platform that can shut you out. A single API or marketplace policy change can end the business.
• Revenue only works at huge scale. Ad-supported or very low price points need audiences small teams rarely reach.
• You are excited about the technology, not the problem. This shows up as high build-fun and low pain scores.
• Competitors exist and you cannot explain why they have not won. Sometimes the answer is that the market does not want it.

If an idea hits two or more of these, treat its score as provisional until you have outside evidence.

How to compare several ideas

Scoring one idea is useful; scoring five side by side is where the framework earns its keep.

1. Score all ideas in one sitting with the same weights, so your standards do not drift.
2. Put them in one table, sorted by total.
3. Look at the spread, not just the winner. If the top three are within a few tenths of each other, the scores are telling you they are roughly equal; choose by founder fit or by which is cheapest to validate.
4. Check each top idea's weakest criterion. A 3.8 with a 1 in reachability can be worse than a 3.5 with no score below 3.
5. Re-score after validation. Interviews and landing page tests should change your numbers. If nothing changes, you probably did not learn much.

Founder fit deserves special attention when comparing. Two ideas with similar totals rarely feel equal to live with for a year. Our guide on choosing a startup idea that fits you covers that side in more depth.

A shortcut: the deep-dive report

Filling in market size, competitors and pricing from scratch takes time. If you want a structured first pass, an Ideas deep-dive report covers most of what this scorecard needs, in these sections:

• Executive summary
• The problem and who has it
• Proposed solution
• Market size and timing, with estimates clearly labeled as estimates
• Competitive landscape, as a table of competitor, positioning and weakness
• Differentiation and moat
• Business model and pricing
• Go-to-market: first 90 days
• MVP scope
• Key risks and mitigations
• Verdict: a frank 1–10 score with reasoning

A report costs 25 credits, and new accounts start with 50 free credits, so you can run it on your top two candidates without paying anything. Treat it as a research assistant rather than an oracle: use it to fill gaps in your scorecard, then check the claims that matter most yourself. More on how reports work is in deep-dive reports and landing pages.

From score to decision

A scorecard is a tool for thinking clearly, not a replacement for evidence. Use it to narrow a long list to one or two candidates, identify the riskiest assumption in each, and then go test that assumption with real people. The idea that survives contact with customers is the one worth building, whatever its original score.