Choose a business idea by comparing evidence of a customer problem with your ability to deliver a useful solution. Enthusiasm is helpful, but it cannot tell you whether a buyer has a problem often enough, cares enough to change, or can purchase what you offer. Start with a shortlist of three ideas and investigate the riskiest assumption in each.
Describe a buying situation, not an industry
“A business helping local shops” is too broad to investigate. “Independent shops need someone to photograph new stock before their weekly online update” gives you a customer, a recurring job and a time constraint. You can now ask how shops handle that job, what delays it and who decides whether to pay for help.
Write each candidate in this format: a particular customer faces a particular problem when a particular event happens, and currently responds in a particular way. Leave your proposed solution out of this first sentence. Otherwise, it becomes easy to reinterpret every conversation as support for the thing you already want to build.
Build an evidence ladder
Use the following ladder as an editorial planning tool, not a scientific scoring model. Higher levels usually require more effort from the prospective customer and reveal more about actual behavior.
- Opinion: someone says the idea sounds interesting.
- Specific history: someone describes a recent incident, consequence and workaround.
- Existing effort: the customer spends time or money addressing the problem today.
- Concrete next step: an appropriate buyer agrees to review a defined offer or arrange an assessment.
- Bounded purchase: a customer buys a clearly described pilot that you can responsibly deliver.
A paid pilot is stronger evidence than a compliment, but one sale does not establish a repeatable business. A friend buying to encourage you is different from an unrelated buyer comparing alternatives. Record that distinction instead of combining every positive interaction into a single total.
Compare ideas on the same worksheet
Create columns for the customer, problem frequency, consequence, current alternative, decision maker, evidence collected and next unknown. Add delivery requirements and the largest commitment needed before you could test the idea. Use plain observations rather than unsupported market-size estimates.
Give each idea a status: investigate, test or park. Investigate means the problem remains unclear. Test means you understand enough to propose a limited next step. Park means the available evidence or delivery demands do not justify more work now. Parking an idea is a resource decision, not a judgment about your creativity.
Hypothetical example: comparing three opportunities
Suppose a founder considers stock photography for shops, custom printed stationery and general administrative support. These are illustrative options, with no assumed local demand. Interviews reveal that two shops repeatedly postpone product updates because photographs are missing. Stationery conversations produce compliments but no recent purchasing examples. Administrative work is real, but buyers want daytime availability the founder cannot provide.
The founder selects photography for a small test, keeps stationery in research and parks administrative support. The decision does not prove photography will succeed. It identifies a testable problem that fits current constraints. The next experiment could be one paid session with a defined image count, turnaround and revision allowance.
Set a learning budget before testing
Choose a limit for research hours, spending and the number of experiments you can realistically support. State the decision each experiment should unlock. A landing page that only counts visits may tell you little if your key uncertainty is whether shop owners can prepare stock for a scheduled session.
Keep evidence that contradicts your idea. Record customers who do not experience the problem, those satisfied with existing options and those unable to authorize a purchase. These findings help narrow the first offer. They also prevent an enthusiastic minority from becoming an imaginary entire market.
Decision checklist
- Can you name the first customer group without using “everyone”?
- Do you have recent examples of the problem rather than hypothetical enthusiasm?
- Can you explain what customers already do instead?
- Have you separated users, buyers and friendly supporters?
- Can you test the idea without a large irreversible commitment?
- Have you written a result that would make you revise or park it?
The SBA planning guidance and the Australian Government research guide provide background on investigating customers and alternatives. The comparison worksheet above is an original planning exercise; its usefulness comes from recording observable evidence and making the next decision explicit.
