Why Did My Business Idea Test Give Me No Clear Answer?
Short answer: Your test was probably unclear because it mixed customer types, changed the offer, measured low-commitment interest or had no pass rule set in advance.
Do not collect more of the same evidence. Rebuild the test around one segment, one result, one price and one customer action for 10 to 20 qualified prospects, then decide against a written threshold. Change only one variable if you run a second round.
Mixed feedback does not necessarily mean the market is divided. It may mean you asked different people different questions under different conditions and combined their answers.
An inconclusive test is usually a test-design result, not a verdict on the idea. Your first job is to find which part prevented interpretation. Extending a broken test only creates a larger pile of ambiguous evidence.
Use the Five-Lock Test Resolver
The Five-Lock Test Resolver checks whether the result can support a decision. Each lock must be closed before you interpret response rates.
| Lock | Test question | A closed lock looks like |
| Segment | Did all participants share the same relevant problem and buying role? | One defined buyer group with recorded qualification |
| Offer | One written offer held constant | |
| Exposure | Did enough suitable people actually receive and understand it? | A known denominator of qualified approaches |
| Sacrifice | Payment, deposit, paid trial or another meaningful commitment | |
| Rule | Did you define pass, revise and stop outcomes beforehand? | A written threshold applied without moving it |
If one lock is open, diagnose it before changing the business. A test with five customer types and no price cannot be repaired by surveying another 100 people.
Separate contradictory evidence from incomparable evidence
Contradictory evidence occurs when similar qualified buyers respond differently to the same offer. That can reveal real variation in urgency, trust or timing.
Incomparable evidence occurs when one person sees a free trial, another sees a £500 package and a third comments on a general description. Combining them produces a percentage with no commercial meaning.
Create a row for every participant showing segment, problem event, offer seen, requested action and response.
If the rows differ in more than one column, split them before analysing.
My view is that founders should discard incomparable responses rather than rescue them with interpretation. Time spent collecting evidence is already gone. Treating weak data as valuable because it was hard to obtain creates a second loss.
Check whether the action was strong enough
Likes, survey answers and compliments can test comprehension. They rarely test purchase. State what the action required the customer to sacrifice.
| Action | Useful for | Weakness |
| General opinion | Language and obvious objections | No personal consequence |
| Contact detail | Very low commitment | |
| Relevant information or booked call | Problem relevance and effort | May still avoid price |
| Deposit or paid pilot | Commercial intent under stated terms | Creates fulfilment responsibility |
A weak action can produce a clear answer to a narrow question. Twenty people correctly understanding the offer proves comprehension. It does not prove demand, and the test is not inconclusive merely because it cannot answer a question it was never designed to test.
Restore the missing denominator
“Five people were interested” is uninterpretable without knowing who was exposed. Five responses from eight qualified buyers differ from five responses after 1,000 broad views.
Record approaches that reached a suitable person, not every impression. For direct research, qualification might mean the prospect has the workflow, volume and buying role specified. For local tests, count people inside the relevant catchment and opening period.
Sample size depends on price, market and buying cycle. Ten to 20 qualified prospects is a practical diagnostic round for a simple early offer, not proof of a market. A high-value sale may justify fewer, deeper decisions; a low-cost consumer product needs more behavioural evidence.
Set decision rules before response arrives
Write three possible outcomes:
- Pass: the required number accepts commercial terms and contribution can work.
- Revise: one repeated, fixable objection blocks otherwise qualified buyers.
- Stop: value is below viable price, the buyer is unreachable or delivery cannot meet the promise.
Avoid a rule based only on percentage interest. Include economics. Three paid customers can be a failure if each job loses £100.
Do not lower the pass threshold after a near miss. Record the result, then choose a deliberately different test.
Moving the rule destroys the main benefit of setting it.
Worked example: Sofia's price-file cleaning service
Sofia tests a service that cleans supplier price files for independent catering businesses. In the first round, she approaches cafés, event caterers and wholesalers. Some see a free review, others hear prices from £150 to £350, and several are asked only whether the idea sounds useful.
She receives 12 positive comments from 30 contacts but no interpretable result. The segment, offer and sacrifice locks are open.
Sofia reruns the test with 20 independent event caterers that update several supplier files each month. Every prospect sees the same offer: one cleaned and reconciled file, delivered within five working days for £260. She asks for a paid booking.
Three buy. Each job takes six hours. Sofia values her time at £30 an hour, and secure transfer and customer-specific delivery costs are £20 per job.
Economic surplus per job is:
£260 minus (6 × £30) minus £20 = £60.
Across three jobs:
- Revenue: 3 × £260 = £780
- Direct delivery costs: 3 × £20 = £60
- Time value: 18 × £30 = £540
- Economic surplus: £780 minus £60 minus £540 = £180
The second test gives a clear, modest pass: qualified buyers paid and delivery produced positive surplus. It does not prove recurring demand. Sofia's next test should hold segment and price constant while asking whether the same customers pay for a second file or monthly cycle.
Change one variable in the next round
When a test misses, select the earliest failed lock. If qualified buyers consistently report insufficient value, change the problem or segment, not the colour of the offer. If value is clear but approval fails, test the budget owner.
Keep a brief record:
- What remained constant
- The one variable changed
- Why the evidence justified that change
- The new pass and stop rules
Multiple simultaneous changes may produce sales, but they will not tell you why. That matters when you try to repeat the result.
Where payment, customer data, regulated advice or consumer promises are involved, ensure the test terms meet local requirements. Rules vary by country and sector, so obtain qualified local advice for material legal, tax, safety or privacy decisions.
Rerun the test in ten days
Spend day one reconstructing every response with the Five-Lock Test Resolver. Discard incomparable evidence and identify the first open lock. By day three, write one offer and qualify a fresh list.
Then act in this order:
- Set pass, revise and stop thresholds before contact.
- Present the unchanged offer to 10 to 20 suitable prospects.
- Request one commercial action and record the denominator.
- Calculate contribution and founder time for any sale.
- Make the decision on day ten without extending the test to avoid an answer.
A clear no protects you from further spending. That is a successful test result.
Frequently asked questions
Does an inconclusive test mean there is no demand?
No. It means the evidence cannot distinguish demand from the way the test was run. Check segment, offer, exposure, sacrifice and rule before judging the market. If all five locks were closed and qualified buyers still avoided a meaningful commitment, weak demand becomes a reasonable conclusion.
If price or segment varied, rerun a controlled test. The exception is decisive external evidence, such as a legal barrier or delivery cost above any credible customer value. That can stop the idea even when customer response itself remains unclear.
How many people should be in a business idea test?
Use enough qualified decisions to reveal repeated mechanisms, not a universal survey number. For a simple early offer, 10 to 20 well-qualified prospects can support a diagnostic round. High-value B2B decisions may need fewer but deeper buying conversations, while low-cost consumer offers need more observed behaviour.
Record the denominator and qualification. A large mixed sample is weaker than a smaller consistent one. The exception is a safety-critical or capital-heavy idea, where customer interest alone is insufficient and technical, legal and financial evidence must also meet a higher standard.
What should I do with evenly split feedback?
Check whether the two groups differ by trigger, volume, role or current alternative. A 50-50 result may reveal two segments rather than uncertainty inside one. Compare who paid or accepted the strongest action, not who expressed preference.
If similar qualified buyers genuinely split under identical terms, test one suspected variable in a second round. Do not average incompatible needs into a broader offer. The exception is a product designed around personal taste, where variation may be expected. You still need to identify whether enough buyers choose one version at a viable price.
Should I test a different price after unclear results?
Only when the original price was shown consistently and qualified buyers identified it as the remaining barrier. If some saw no price, you have not tested the first price properly. Hold scope and segment constant, choose one new price and calculate whether it still supports contribution.
A lower price that wins customers while creating losses is a clear failure, not progress. The exception is a smaller pilot whose reduced price reflects genuinely reduced delivery. State the future price so you do not mistake pilot-only demand for demand at normal terms.
Can I reuse the same prospects for a second test?
Yes, when the change is material, you explain it and repeated contact will not distort behaviour. Existing prospects can compare the revised offer, but they may also answer helpfully because they know you. Add fresh qualified buyers so you can separate learning effects from commercial response.
Do not repeatedly ask the same people to validate minor changes. The exception is a service built around ongoing customer development, where existing buyer participation is part of delivery. Even then, payment and use should carry more weight than their stated encouragement.
When should I stop repeating tests?
Stop when a fundamental condition repeatedly fails, the economics cannot support a viable price, or each round changes the idea without preserving evidence. Set a maximum number of controlled revisions before starting, such as one for buyer and one for offer. You do not owe an idea endless chances.
Continue when a specific change follows repeated evidence and the next test can answer it cheaply. The exception is a long buying cycle where no decision was realistically due during the test period. In that case, redesign the milestone rather than declaring either success or failure prematurely.
Comments (0)