From Idea to Sustainable Enterprise · 5 of 10The Enterprise Beneath the Idea
A Good Test Should Put the Assumption at Risk
Why validation becomes meaningless when an experiment is designed only to confirm what the founder already wants to believe.
A founder has developed an idea for a new service.
They describe it to friends, former colleagues and several people in their professional network.
The response is encouraging.
People say the idea is interesting. Some describe it as necessary. Others say they would consider using it or know someone who might benefit.
The founder creates a social-media post announcing that the service is coming soon.
It receives likes, supportive comments and several requests for more information.
The idea appears to have been validated.
But no one has been asked to make a decision.
Nobody has seen the complete offer, compared it with an alternative, committed time, shared information, paid a deposit or used the service in a real situation.
The founder has received encouragement.
They have not yet obtained strong evidence.
This distinction is easily lost because positive reactions feel like progress. They reduce uncertainty and strengthen the founder’s emotional commitment to the idea.
But validation is not the process of collecting reasons to continue.
It is the process of creating evidence that could justify continuing, changing, pausing or stopping.
A good test should put the assumption at risk.
If an experiment can only confirm what the founder already believes, it is not a meaningful test.
It is reassurance presented as research.
Every idea contains assumptions
An emerging venture cannot begin with complete knowledge.
The founder must make provisional judgements about the problem, audience, value, price, delivery model and route to market.
They may believe that:
A particular group experiences the problem
The problem is important enough to motivate action
Existing alternatives are inadequate
People will understand the proposed offer
Someone will be willing to pay
The service can be delivered at a sustainable cost
A trusted route to the audience exists
The founder has the necessary capability
The technology will function as expected
Participants will accept the way their information is used
A funder or partner will recognise the intended value
These are not signs of poor planning.
They are the unavoidable foundations on which an unproven idea initially rests.
The danger appears when assumptions become embedded in the venture without remaining visible.
A belief enters the business plan, is repeated in presentations and eventually becomes treated as a fact. Decisions accumulate around it. The cost of questioning it increases.
Testing helps prevent this.
But only if the founder knows exactly which assumption the test is intended to examine.
“Do people like the idea?” is usually the wrong question
People are often generous when responding to a new idea.
They want to encourage the founder. They may find the general concept admirable or interesting. Saying that they dislike it can feel unnecessarily negative, particularly when the relationship matters.
A question such as “Would you use this?” asks someone to predict future behaviour in an imagined situation.
There is little cost to saying yes.
The person does not have to compare the service with other priorities, examine the price, find time, obtain approval or accept the practical conditions of participation.
This is why stated enthusiasm frequently fails to become action.
The problem is not necessarily dishonesty.
A person can sincerely like an idea while deciding not to use it.
They may recognise its value but not experience the problem themselves. They may want the outcome but not enough to change their current behaviour. They may support the purpose but find the price, timing or process unsuitable.
A stronger enquiry focuses on experience and behaviour:
When did you last encounter this problem?
What did you do?
What did that cost in time, money or effort?
What alternatives did you consider?
Why did you choose the current option?
What made you continue or abandon the process?
Who else influenced the decision?
What would have to change for you to try something different?
These questions do not ask the person to endorse the founder.
They help the founder understand the situation into which the idea would enter.
Test the most dangerous belief first
Not every assumption deserves equal attention.
A founder may spend weeks testing the colour, name or format of an offer while leaving its central commercial or cultural assumption untouched.
The useful starting point is the belief that would cause the greatest damage if it were false.
This may be the assumption that the problem exists at sufficient scale.
It may concern willingness to pay, access to customers, the cost of delivery or the founder’s right to intervene.
For a community project, the most dangerous belief might be that intended participants trust the organisation.
For an AI-enabled service, it might be that people are comfortable sharing the information required by the system.
For a creative product, it might be that the audience values the distinctive quality that makes production expensive.
For a consultancy, it might be that potential clients recognise the problem before they are already committed to the wrong course of action.
Testing smaller details may feel safer.
It does not reduce the principal risk.
A test should be connected to a decision
An experiment creates value when its result can change what happens next.
Before running a test, the founder should be able to complete three statements:
We currently believe…
We will examine this by…
Depending on the result, we will…
The final statement is the most important.
If the founder cannot explain how different results would affect the venture, the test may be too vague.
Suppose a founder believes that independent creative practitioners would pay for a structured review before developing a major project.
They offer five paid pilot places at a reduced but meaningful price.
Several outcomes are possible:
All places are purchased quickly.
People enquire but do not purchase.
People purchase but need extensive explanation first.
Participants value the review but say the timing was wrong.
Participants complete the process but do not use the recommendations.
Practitioners express interest, but organisations are more willing to pay.
The cost of delivering the pilot is much higher than expected.
Each outcome provides different intelligence.
The founder needs to decide in advance what would justify continuation, revision or further investigation.
Without this discipline, almost any result can be interpreted positively.
No sales become evidence that the marketing needs improvement.
Low participation becomes evidence that more promotion is required.
Negative feedback becomes proof that the audience was wrong.
The idea survives every test because the founder changes the explanation instead of the venture.
That is not validation.
It is protection.
The smallest test is not always the cheapest test
Startup advice often recommends building a minimum viable product.
The principle is useful: do not invest in a complete venture before investigating its most important assumptions.
But “minimum” can be misunderstood.
A test that is too small, incomplete or poorly delivered may fail because it never gives people a credible experience of the proposed value.
A premium creative service cannot be tested fairly through an unprofessional document that undermines the intended positioning.
A relationship-based community programme cannot be reduced to an online survey and expected to reveal how trust would develop.
A strategic service cannot be judged through a brief conversation if its value depends on careful analysis and a considered written output.
The objective is not to create the cheapest possible version.
It is to create the smallest credible version capable of answering the question.
Credibility may require an appropriate standard of design, accessibility, care or human interaction.
Anything less may test the weakness of the prototype rather than the strength of the idea.
Protect the character of the idea
Testing can also distort a venture when the easiest things to measure become the only things preserved.
A founder may remove cultural depth because a simpler offer is easier to explain.
A community initiative may narrow participation to people who are easiest to recruit.
An artist may produce more conventional work because unfamiliar ideas receive slower responses.
A service based on careful human attention may be automated to reduce the cost of a pilot.
These changes might improve short-term performance while destroying the qualities that made the idea distinctive.
Before testing begins, the founder should identify the protected principles established in the Origin Map.
These might include:
Cultural integrity
Community influence
Accessibility
Creative independence
Environmental responsibility
Human judgement
Fair payment
Privacy
A refusal to use manipulative marketing
Respect for participants’ right not to use AI
The protected principle is not the same as a protected solution.
A founder may remain committed to community ownership while changing the service format.
They may protect creative independence while revising the commercial model.
They may retain human accountability while using technology for appropriate administrative work.
Testing should place assumptions at risk.
It should not automatically place the venture’s ethical foundation at risk.
A pilot is not merely a discounted launch
Organisations often describe the first public version of an offer as a pilot.
But a pilot is useful only if it has a defined learning purpose.
A discounted launch aims primarily to attract early customers.
A pilot aims to investigate how the offer works in practice.
The two can overlap, but the distinction should remain clear.
A credible pilot establishes:
What is being tested
Why the selected participants are relevant
What remains unfinished
What evidence will be collected
How participant information will be used
What participants will receive
Which limitations are already known
What decisions will follow
What support is available if something goes wrong
Participants should not unknowingly absorb the risks of an underdeveloped service.
If the founder needs substantial feedback, additional time or tolerance for change, the exchange should recognise that contribution.
A lower price may be appropriate.
In some circumstances, participants should be paid for their involvement rather than asked to pay for access.
The correct arrangement depends on who receives value and who performs the work.
Free participation can produce misleading evidence
Free pilots are attractive because they reduce the barrier to participation.
They can help examine usability, relevance, delivery and participant experience.
But they provide weak evidence about willingness to pay.
Removing the price changes the decision.
A person may attend a free workshop because it sounds interesting. The same person may not purchase it when it competes with other demands on their income.
Free participation can also create lower commitment. People register and do not attend, delay completing exercises or give less attention to the process because little has been risked.
This does not make free testing useless.
It means the founder should be clear about what it can establish.
A free pilot might test:
Whether people understand the activity
Whether the format is accessible
Where confusion appears
Whether the proposed outcome occurs
How long delivery takes
What support participants require
It cannot, by itself, confirm a sustainable commercial model.
Different assumptions require different tests.
Payment is evidence—but it is not complete proof
A purchase is stronger evidence than a compliment.
Someone has made a real decision and accepted a financial cost.
But a small number of sales does not confirm every part of the venture.
The customer may have purchased because they trust the founder personally.
The introductory price may not cover sustainable delivery.
The first buyers may come from an existing network that cannot be expanded.
People may purchase once but see no reason to return.
The service may create value for customers while exhausting the founder.
A successful paid pilot therefore creates new questions:
Why did people buy?
What nearly prevented the purchase?
Did the experience fulfil the promise?
What part created the greatest value?
What was unused or unnecessary?
Would they pay the standard price?
Would they recommend it?
Can the founder deliver it repeatedly?
How much unpaid preparation did the pilot require?
Does the model remain viable without personal relationships?
Payment is consequential evidence.
It is not the end of learning.
Test behaviour in context
People make decisions within systems.
A customer’s response to an offer is affected by timing, income, organisational approval, relationships, competing priorities and perceived risk.
A test should therefore resemble the context in which the real decision will occur.
If an organisation must approve the purchase, speaking only with the intended participant is insufficient.
If the service depends on a trusted community partner, testing direct advertising may answer the wrong question.
If customers normally discover the problem only after a triggering event, promoting the offer to a broad audience may produce weak interest even when the service has real value.
The founder needs to understand the complete decision journey:
1. What causes someone to recognise the need?
2. What language do they use?
3. Where do they look for help?
4. Who or what do they trust?
5. What alternatives do they consider?
6. Who controls the budget?
7. What risks concern them?
8. What evidence reduces uncertainty?
9. What finally produces action?
10. What happens after the purchase?
Testing one isolated message cannot validate the entire journey.
Ethical testing requires informed participation
Experiments involving people create responsibilities.
A founder may collect personal stories, observe behaviour, test different prices or use AI to analyse responses. These activities can create value for the venture, but they may also introduce privacy, consent and power concerns.
Participants should understand, at an appropriate level:
That the offer is being tested
What information is being collected
How the information will be used
Whether responses will be identifiable
Who will have access
Whether AI tools are involved
What participation requires
How they can withdraw
What will happen after the pilot
Whether their contribution may shape a commercial product
The formality required will depend on the level of risk.
Testing a new website headline does not require the same safeguards as collecting accounts of trauma, financial insecurity or discrimination.
The principle is proportionality.
The greater the sensitivity, vulnerability or consequence, the stronger the protection should be.
“Move fast” is not a defence for careless treatment of people.
Communities should not become permanent test environments
Community and cultural organisations are frequently invited to participate in pilots.
A new initiative arrives. Residents share experiences, attend workshops and help refine the model. Funding ends, the organisation leaves and another pilot begins later.
The community repeatedly contributes knowledge without receiving stable infrastructure or long-term benefit.
This creates consultation fatigue and institutional distrust.
A culturally intelligent founder should investigate previous activity before designing another experiment.
Ask:
Has something similar already been tested?
What happened to the findings?
Did participants see any result?
Which organisations already hold relevant knowledge?
Could the venture support or extend existing work?
Is another pilot necessary?
What continuing value will remain after the test ends?
Innovation does not always require creating something new.
It may require recognising, connecting or sustaining what already works.
Negative results are not failed tests
A test may show that people do not recognise the problem, will not pay the proposed price or prefer an existing alternative.
This can be disappointing.
But the test has performed its function.
A negative result obtained early can prevent larger losses later. It may protect the founder’s money, time and reputation. It may prevent communities or partners from being asked to support an unsuitable intervention.
The founder’s task is to interpret the result carefully.
A poor response could mean:
The problem is not important
The selected audience is wrong
The timing is unsuitable
The offer is unclear
The route to market is weak
The price is inappropriate
The test did not provide a credible experience
Trust has not been established
A better alternative already exists
The solution does not address the real problem
Further investigation may be required before choosing among these explanations.
But disappointment should not be converted automatically into a marketing problem.
Sometimes the market does understand the offer.
It simply does not want it enough.
That is important intelligence.
Success criteria must be defined before the result
Founders can protect themselves from convenient interpretation by defining thresholds in advance.
For example:
We will offer ten paid pilot places at £75. If at least six are purchased by people outside our immediate personal network, we will proceed to test delivery and standard pricing. If fewer than three are purchased, we will return to the problem and value proposition before investing in development.
The exact numbers are not universal. They should reflect the venture, price, audience and purpose of the test.
The important principle is that the founder decides what the result means before knowing what the result is.
Qualitative tests also need criteria.
A community workshop might examine whether participants believe the problem has been described accurately. The founder could decide that substantial disagreement from several differently positioned participants will trigger a revision.
A prototype might test whether users can complete a task without assistance.
A service pilot might examine whether participants use the final output to make a real decision within four weeks.
Clear criteria turn activity into evidence.
Do not allow metrics to erase meaning
Quantitative measures are useful.
Conversion, completion, retention, cost and time can reveal patterns that intuition misses.
But a metric can conceal important differences.
A workshop may achieve high attendance while participants feel unable to influence the discussion.
A digital service may increase completion by removing necessary opportunities for human judgement.
A community programme may meet its target by repeatedly engaging the same confident participants.
A marketing campaign may produce enquiries that are poorly matched to the service.
The founder should examine both performance and experience.
Ask:
What happened?
For whom?
Under what conditions?
Who was excluded?
What did the number fail to show?
Did the result strengthen the purpose or merely the metric?
Evidence must remain connected to meaning.
Build an Experiment and Learning Map
At this stage, the founder can create an Experiment and Learning Map.
The map should include:
The assumption
What does the venture currently believe?
The consequence
What would happen if that belief were false?
Existing evidence
What is already known?
The question
What precise uncertainty should the experiment reduce?
The method
What is the smallest credible way to investigate it?
Participants
Whose behaviour, knowledge or experience is relevant?
Ethical conditions
What consent, privacy, payment or protection is required?
Success threshold
What result would support continuation?
Warning threshold
What result would require revision or further research?
Stop condition
What result would make continued investment irresponsible?
Learning
What happened—and what might explain it?
Decision
What changes as a result?
The map creates a visible chain between belief, evidence and action.
It prevents experiments from becoming disconnected activities whose results are filed away but never allowed to affect the venture.
A practical field exercise
Choose the most consequential assumption from the Living Intelligence Record.
Complete the following:
We believe
Write one specific statement.
For example:
We believe early-stage creative founders will pay for an independent review before investing in a full business plan.
Because
List the evidence currently supporting the belief.
Separate direct evidence from interpretation.
We need to learn
Identify the remaining uncertainty.
We will test it by
Design the smallest credible experiment.
We will measure
Choose the behaviours, outcomes and experiences that matter.
We will protect
Record the ethical and cultural principles that must remain intact.
We will continue if
Set the success threshold.
We will revise if
Define the result requiring change.
We will stop if
Identify the evidence that would make further investment inappropriate.
We will decide by
Set a clear date for reviewing the evidence.
A test without a decision date can continue indefinitely while the founder waits for a more favourable result.
The decision to record
At the end of this stage, the founder should record:
The most consequential assumption
Why it matters
Current supporting evidence
Contradictory evidence
The experiment design
Relevant participants
Ethical safeguards
Measures and evidence thresholds
The result
Alternative explanations
The resulting decision
Changes made to the idea
Questions requiring further investigation
The most important entry is not whether the test “passed.”
It is:
What do we now believe, and what will we do differently because of what we learned?
Validation should make the idea more vulnerable—and more credible
Founders often approach testing as if the objective were to defend the idea.
They want evidence that will make them feel confident, persuade an investor or justify the next stage of development.
But an idea becomes stronger when it is allowed to encounter reality before the cost of change becomes too high.
A useful experiment makes the venture temporarily vulnerable.
The audience may not respond.
The value proposition may fail.
The price may prove unrealistic.
The community may reject the founder’s interpretation.
The delivery model may require more time and care than the business can afford.
These are difficult discoveries.
They are also precisely what testing is for.
The alternative is to protect the idea from meaningful challenge until it becomes a fully developed venture built on assumptions nobody was willing to risk.
A good test does not ask the world to confirm the founder’s confidence.
It creates a fair opportunity for the evidence to disagree.
If the idea survives, it does so with greater credibility.
If it changes, it becomes more responsive.
If it stops, the founder has prevented further resources from following a false assumption.
In each case, the test has produced something valuable.
Not certainty.
Better judgement.
Learning Path Reflection
Before continuing, consider:
1. What is the most consequential assumption behind your idea?
2. What evidence currently supports it?
3. What evidence might contradict it?
4. Have you asked people to act—or only to express an opinion?
5. What is the smallest credible version of the offer?
6. What must be protected while the idea is tested?
7. Who will contribute knowledge, time or risk?
8. What safeguards or compensation are appropriate?
9. What result would cause you to revise the idea?
10. What result would cause you to stop?
Living Intelligence Record
Record:
The assumption being tested
Its potential consequence
Existing evidence
The experiment question
The test design
Participants and recruitment route
Ethical and cultural safeguards
Success, warning and stop thresholds
Results
Contradictory evidence
Interpretations
The decision made
Changes to the idea
The next uncertainty
Related Map
Experiment and Learning Map
Continue the Learning Path
Next article: The Wrong Business Model Can Betray the Right Idea
The next stage examines how revenue, pricing, funding, ownership and costs can support an enterprise’s purpose—or gradually redirect it towards the priorities of whoever controls the money.
About This Series
This article is part of The Enterprise Beneath the Idea, the original Cultural Intelligence Studio article collection accompanying the From Idea to Sustainable Enterprise learning path.
The learning path combines original CIS thinking with carefully selected videos, podcast conversations, practical exercises, a Living Intelligence Record and connected cultural intelligence maps.
Its purpose is to help people make stronger decisions about what should be developed, changed, tested, paused or left behind.
Optional CIS Support
The Idea Clarity and Validation Review provides an independent examination of an emerging idea, its intended audience, underlying assumptions, proposed value, available evidence and smallest credible test.
The review is designed to improve the decision—not to guarantee that the original idea should proceed.
Engaging CIS is optional. The most responsible next step may be a small independent experiment rather than additional paid development support.