Free field notes / Updated 2026-10-21

How to Choose Between Business Ideas

Compare business ideas with a four-dimension rubric, a worked scoring example, and separate risk checks before choosing one small test to run.

Choose between business ideas by comparing how clear the problem is, how easily you can reach buyers, how little a real test costs, and how quickly it can produce a useful signal. Score the ideas using the same definitions and keep risk flags separate. The result helps you choose what to test next, not predict which business will succeed.

Turn a long list into comparable statements

Ideas can come from your own experience, an idea database, customer conversations, or an AI brainstorm. Their source does not establish demand. Treat every candidate as a claim that still needs evidence.

Rewrite each idea into the same form: "For this person, in this situation, I would deliver this outcome." Replace vague categories such as "productivity" with a task someone actually performs.

For example, "help hobby clubs" becomes "help volunteer club organizers turn last month's meeting notes into a clean event task list." The clearer sentence makes buyer access, delivery, and test costs easier to examine.

Start with three to five candidates if the list is long. That is a practical rule of thumb for keeping the exercise manageable, not a claim that a certain list length produces better businesses.

Screen risk flags before assigning scores

Regulated finance, medical, and legal ideas stop here. Napkinproof does not give them a score or a validation plan. Seek qualified guidance before testing with real people.

For other ideas, record risk flags separately rather than subtracting a few points and moving on:

• Safety: physical harm, children, food handling, or vehicles can require further checks before a test.
• Legal-exposure-general: permits, licenses, intellectual property, or other people's data may require appropriate review.
• Platform dependence: the idea may rely on another service's access or rules.
• Capital-heavy: even the first meaningful step may require money you cannot readily risk.
• Ethics: deception, surveillance, gambling, or adult activity can make an ordinary test inappropriate.

A high total cannot cancel a flag. Keep a one-sentence note explaining what must be resolved. For regulated ideas, use the general starting-point guide rather than continuing the scoring exercise.

Use the four dimensions exactly

Each dimension is 1 to 5, higher is better:

• Problem clarity: 5 = a specific person with a specific, frequent, painful problem they already spend time or money on. 1 = vague "people would like this".
• Buyer reachability: 5 = you can name where 20 likely buyers gather and reach them this week for free. 1 = buyers are hard to find or only reachable through paid ads or enterprise sales.
• Cost to test: 5 = a real test costs under US$20. 1 = needs inventory, licenses, hardware or over US$500.
• Speed to signal: 5 = a meaningful yes/no signal is possible within 48 hours. 1 = needs months.

Overall: sum of four (4 to 20). 16 to 20 "Test this weekend", 11 to 15 "Sharpen first, then test", 4 to 10 "Rethink before testing".

Use intermediate values of 2, 3, or 4 when the evidence sits between the stated endpoints. Write a short reason for the choice. The rubric does not define precise formulas for those intermediate values, so avoid treating your judgment as a measured probability.

The cost dimension is about a real test, not the price of writing a description. If the claim requires a working physical mechanism, a free sketch does not automatically earn a 5 for cost.

Score three made-up ideas

The following table uses hypothetical circumstances. The scores describe this imagined founder's access and evidence, not the inherent value of each idea.

For the first idea, imagine the founder has helped two clubs prepare events, has permission to contact a group of twenty organizers, and can deliver a manual sample with existing equipment. The founder has seen a specific task but not enough evidence of painful recurring effort for a 5 in problem clarity.

For the puzzle-night worksheets, the founder knows a few hosts but cannot yet name a free route to twenty likely buyers. Materials are inexpensive, but the next actual event is two weeks away. That slows a meaningful usage signal even if people can comment on the concept today.

For the storage unit, imagine the founder's test requires materials, hardware, and a functional outdoor prototype costing over US$500. Fabrication and realistic use observations would take months. It also carries safety and capital-heavy flags that remain separate from its score.

Choose the next action from the result

1. For 19: Run a small weekend test of the task-list service, provided relevant flags are addressed.
2. For 14: Sharpen the worksheet audience and identify a nearer use occasion before testing.
3. For 7: Rethink whether a lower-cost, meaningful first question exists before committing resources.

Do not add all three scores together or average away the weak dimensions. Each row represents a separate candidate. A tie is a reason to examine your evidence, not to invent extra decimal places.

If two ideas remain close, consider your access and ability to deliver the test. Personal interest matters because you must do the work, but keep that preference separate from claims about buyer demand.

Check what the score does not tell you

A 19 does not mean a 95% chance of success. The total is a triage aid, not a probability, market-size estimate, valuation, or earnings forecast.

The rubric favors ideas that are cheap and fast to test. That is appropriate for choosing a first experiment with little money. It does not mean slower or more expensive ideas are inherently worthless.

Likewise, a low score may reflect your present circumstances. Someone else may already have access, equipment, or evidence you lack. Score your actual position today rather than an imagined future version of yourself.

After selecting a candidate, map its riskiest assumption. The idea score tells you which row deserves attention. The assumption map tells you what the next test should challenge.

Common mistakes

• Giving high scores because an idea came from an impressive source.
• Treating a free concept image as a complete physical-product test.
• Counting risk flags inside the total instead of reviewing them separately.
• Scoring a regulated idea after labeling it risky.
• Adjusting intermediate values to make a favorite idea win.
• Using the total as a claim about likely revenue.

Keep the original score sheet and date it. After a test, update the reasons as well as the numbers so you can see which observations changed your view.

Next step

Paste your idea into the free triage tool for a consistent first pass, then compare it with two other candidates using the same definitions. Choose one modest test rather than trying to build every promising idea at once.