To choose which product idea to pursue first, filter them against hard disqualifiers, score the survivors on the same criteria side by side, then spend a small amount of money testing the top two in the real world. Most people skip straight to the one that excites them most, which is how promising ideas become expensive lessons. A structured comparison takes an evening and repeatedly changes the answer — the idea that looks best on paper is often not the one that survives contact with buyers.

The problem: they all look promising

Ideas are seductive precisely because they are unexamined. Each one exists in your head in its best possible version, with no cost, no competitor, and no manufacturing constraint. Comparing them fairly means forcing all of them to the same level of detail and judging them on criteria you fixed in advance, before enthusiasm votes.

Step one: the coarse filter

Write each idea in two sentences: who has the problem, and what the product does about it. Then eliminate on hard grounds:

  • Already exists and is done well. Search retail, marketplaces, and patents before anything else. Our list of free patent search tools covers the free options, and half an hour there saves months.
  • You cannot describe the buyer. If the answer is "everyone," the idea is not ready.
  • The regulatory path exceeds your resources. Not all approvals are equal — a medical or children's product carries a burden a desk accessory does not. Check with does your product need regulatory approval.
  • The unit economics obviously cannot close. If a rough estimate of manufacturing cost lands above the price the market clearly bears, no amount of clever engineering fixes it.
  • You have no honest interest in the field. This one sounds soft and is not. You will spend years in this market.

Two or three ideas usually survive. That is the point.

Step two: a scoring table

Build a simple table: ideas as rows, criteria as columns, one to five per cell, with weights you set before scoring. Useful criteria:

  • Pain intensity — is this an irritation or a real cost? Weight this heavily.
  • Buyer reachability — can you find and sell to these people affordably?
  • Development cost and complexity — mechanical only, or mechanical plus electronics plus firmware plus an app?
  • Time to first revenue.
  • Defensibility — patent potential, know-how, brand, or supply relationships.
  • Margin potential at realistic volume.
  • Your own advantage — domain expertise, industry contacts, distribution you already have. This is the most underrated column on the table.

Score each cell with a note on why. The number is less useful than the argument it forces. If two ideas tie, prefer the one that is cheaper and faster to test — option value beats theoretical upside when your budget is finite. Our guide to running a feasibility study covers how to firm up the engineering columns before you score them.

Step three: a cheap field test

Take the top two and spend a small, capped amount on evidence rather than analysis:

  • Talk to ten people who have the problem. Ask what they do today and what it costs them — never "would you buy this." See how to validate a product idea for question design.
  • Put up a one-page site for each with a clear price and a preorder or waitlist button, and send a small amount of traffic to both. A landing page preorder test compares demand for two ideas in the same week for very little money.
  • Get quotes. A rough manufacturing estimate from a supplier turns your cost guess into a number.
  • Show a crude mockup. Reactions to an object are more informative than reactions to a description; getting honest feedback on a product idea explains how to ask without steering the answer.

Set the decision rule before you start: whichever idea produces more real signal by a fixed date wins. Otherwise you will keep gathering data to avoid choosing.

Traps in the selection process

  • Choosing the most technically interesting idea. Engineering fascination is not demand.
  • Choosing the cheapest to build. Easy to build usually means easy to copy.
  • Pursuing several at once. Two half-developed products lose to one finished one. Divided attention is the most common cause of nothing shipping.
  • Sunk-cost loyalty. The idea you have thought about longest is not the best one; it is just the oldest.
  • Asking only friends and family. They answer the question "do you like me."
  • Waiting for certainty. No amount of research removes the risk. The goal is a defensible choice, not a guaranteed one. Reviewing why new products fail is a useful way to pressure-test your reasoning before committing.

What about the ideas you did not pick?

Do not discard them. Write each one up properly — description, sketches, dates, everything you learned — and file it. Two things happen over time: some become obviously wrong, which is free information, and some become viable because a component got cheap or a regulation changed. Keeping a dated record also protects your position under the first-to-file system; see the inventor's notebook. Revisit the file once a year.

Want a second opinion on your shortlist?

Projects House reviews idea shortlists regularly, and the engineering columns — cost, complexity, and time — are usually where an outside view changes the ranking. Send us your top two or three through the contact form and we will tell you honestly which one we would build first, and why.