Founder guide
How to start a product business when you do not have an idea yet
A method for finding a physical product worth making when you have the intent to build a business but no concept, and for testing candidates cheaply.
Direct answer
Not having an idea is a starting position, not a disqualification. Pick a group of people you already understand, find a purchase they keep making badly, and turn that into a specific product. Then test two or three candidates as real priced offers instead of debating them.
Start from an audience, not from a product
Ideas generated in the abstract are almost always ideas nobody asked for. Ideas generated from a group you already understand come with a distribution channel, a vocabulary, and a way to check whether you are right.
Write down the groups you have genuine access to: a hobby you practice, a profession you worked in, a community you belong to. Access is the scarce resource at the start, and it is the one advantage a first-time founder usually already has and consistently ignores.
Look for repeated bad purchases
The most reliable product opportunities are things people already buy on a repeating basis and are quietly unhappy with. They have proven willingness to pay, an existing budget line, and a complaint they can articulate.
Listen for the workaround: the thing people modify, combine, or tolerate because nothing suitable exists. A workaround is a specification written by a customer who did not know they were writing one.
- Note what people buy repeatedly and complain about anyway
- Note what they modify after buying
- Note what they buy in the wrong size, quantity, or format
- Ignore ideas nobody currently spends money on
Turn an observation into a specific product
A vague observation produces a vague product. "Better gym gear" is not testable. "A chalk container that seals properly for people who train at home and share the space with a carpet" describes something you could make and someone could refuse to buy.
Specificity is what makes a test meaningful. The narrower the product, the clearer the signal, and narrow products are far easier to widen later than broad products are to sharpen.
Test candidates as real offers, not as arguments
Two or three candidate products can be argued about indefinitely or tested in a fortnight. Build each one as a real product with real pricing and a real storefront, put them in front of the audience they were designed for, and let payment decide.
Decide in advance what result would make you continue, revise, or drop each one. A threshold chosen after the numbers arrive is a threshold chosen to protect your favorite.
How to run this in Dough
Describe the audience and the problem rather than the product. Dough generates product concepts from that description, and you refine the ones worth refining until they are specific enough to be wrong.
Build the candidates you want to test, set prices from a visible cost breakdown, attach launch goals, and run each offer to its intended audience. Running several candidates as real priced offers costs a fraction of committing to one on instinct, and the loser teaches you as much as the winner.
What to do with the result
A candidate that converts gives you a product to manufacture and a customer description that already works. A candidate that fails gives you objections in the buyer’s own words, which is the raw material for the next specific idea.
The goal of the first month is not a business. It is to stop guessing, which is the only thing standing between having no idea and having a good one.
What changes about the order
- Getting from an observation to a product
- UsuallyAn idea stays a description until someone is paid to design and build it, so most candidates never get tested.
- With DoughDescribing an audience and a problem produces product concepts you can refine, so a candidate becomes testable the same day.
- Comparing candidates
- UsuallyTwo or three ideas are argued about, and the one with the best story wins rather than the best evidence.
- With DoughEach candidate becomes a real priced offer, and payment decides between them.
- Cost of being wrong
- UsuallyBeing wrong costs a production run and several months.
- With DoughBeing wrong costs a fortnight and gives you objections in the buyer own words.