Most market-validation projects produce information. Far fewer
produce a defensible decision.
The difference usually appears when a team moves from asking people
what they think to testing what they will actually do.
Surveys can reveal language, preferences, stated intentions and broad
attitudes. They can help a business understand how customers describe a
problem. They cannot, on their own, prove that customers will complete
registration, pass verification, fund an account, change an established
habit or return after the first transaction.
That is why market validation is not a survey exercise.
A credible market validation strategy must reduce uncertainty across
four connected questions:
- Is the customer problem important enough to change behaviour?
- Can the business reach the relevant customer at a viable cost?
- Can the product deliver value under local operating conditions?
- Do the resulting economics justify further investment?
If validation answers only the first question, the business still has
a hypothesis.
This distinction has shaped my work across iGaming, digital products
and travel distribution. In each sector, the visible demand signal was
only one part of the decision. The more consequential risks often sat
elsewhere: trust, payment access, identity requirements, partner
quality, fulfilment, customer support and the cost of sustaining the
operating model.
A market may be attractive in theory and commercially inaccessible in
practice.
Stated intent is weak
evidence
People are often sincere when they say they would try a product.
Their answer is still shaped by the artificial conditions of the
question.
The price may be hypothetical. The product may not require them to
upload an identity document. The payment method may not be visible. The
familiar alternative is not competing for attention at that moment.
There is no social risk, delay or uncertainty attached to the
answer.
Real behaviour includes all of those conditions.
This is why responses such as “I would use it,” “the concept is
useful” or “the price seems reasonable” should not be converted directly
into revenue forecasts. They are signals about perception, not proof of
adoption.
The gap becomes wider in markets where trust is uneven, digital
access is fragmented or informal behaviour sits alongside formal
services. A respondent may like the proposition but refuse to pay an
unfamiliar merchant. A customer may register but stop when a preferred
payment method is missing. An agent may praise the product but decline
to promote it when incentives and support requirements become clear.
Validation should therefore create situations in which the customer
has to give something scarce. That may be money, time, data, attention,
reputation or a change in routine.
The more closely the test resembles the real decision, the stronger
the evidence.
Validate the riskiest
assumption first
Teams often begin with the easiest question to research rather than
the assumption most capable of destroying the business case.
They estimate market size, catalogue competitors and run a preference
survey. These activities can be useful. They become a distraction when
the decisive risk is whether customers can deposit, whether a
distribution partner can be trusted, whether unit economics survive
incentives or whether regulation permits the intended route to
market.
A strong validation process starts with an assumption register.
List the conditions that must be true for the proposed market entry
or product to work. Typical assumptions include:
- the problem occurs frequently enough to matter;
- the target customer recognises it;
- the proposed solution is credible;
- customers can access and pay for the service;
- acquisition channels can reach the intended segment;
- local partners can deliver the promised distribution;
- onboarding can be completed under relevant identity rules;
- the business can fulfil the service reliably;
- customer value exceeds the full cost of acquisition and
support; - regulation and advertising rules allow the operating model.
Then rank each assumption by uncertainty and consequence.
An assumption with high uncertainty and high consequence should be
tested before the business invests in polishing lower-risk details. This
sounds obvious, yet many validation programmes protect the original
proposal by choosing tests it is likely to pass.
Validation should expose a weak proposition early. Its purpose is not
to make a launch look justified.
Market size is not
reachable demand
Top-down market estimates are seductive because they create scale
quickly.
A presentation begins with population, smartphone adoption, category
expenditure or total betting turnover. It applies a small percentage and
produces an apparently significant revenue opportunity.
The calculation may be arithmetically correct and commercially
meaningless.
Reachable demand depends on a narrower chain:
- Customers who experience the relevant problem.
- Customers who can be reached through available channels.
- Customers who trust the proposition enough to engage.
- Customers who can complete onboarding and payment.
- Customers whose value supports the cost of service.
Each stage reduces the addressable pool.
The World Bank’s Global Findex and GSMA’s mobile-economy research are
valuable because they help describe access conditions such as account
ownership, mobile connectivity and digital financial behaviour. They
should inform assumptions, not replace transaction-level validation.
Country averages also hide differences by income, gender, age, location
and device quality.
For a commercial decision, the practical question is not “How large
is the market?” It is “How many customers can this specific product
reach, convert and serve under the proposed conditions?”
That question requires bottom-up evidence.
Build an evidence ladder
Not all validation evidence deserves equal weight.
I use an evidence ladder that moves from reported opinion toward
observed commercial behaviour.
Desk research
Regulatory documents, market reports, public statistics, competitor
products and channel data help define the environment. This stage is
essential for framing, but it remains indirect.
Interviews and observation
Customer and partner interviews reveal language, workflow, objections
and existing alternatives. Observation is particularly valuable because
customers often omit routine workarounds when asked to describe their
behaviour.
Commitment signals
Waitlists, letters of intent, partner introductions, completed
qualification forms and willingness to provide data show more commitment
than favourable opinions. They are still not transactions.
Behavioural tests
Landing pages, prototype use, assisted onboarding, test campaigns and
manual service delivery show whether people take the next step under
more realistic conditions.
Commercial evidence
Payment, repeat use, retained activity, fulfilled orders and
contribution after variable cost provide the strongest early
validation.
The appropriate rung depends on the decision. A team exploring a
problem does not need a complete commercial launch. A board approving
substantial market-entry investment should not rely primarily on
interview enthusiasm.
Every conclusion should state the strength and limits of its
evidence.
Test the
customer journey, not only the proposition
A product can solve a real problem and still fail because the
surrounding journey is broken.
In digital acquisition, the visible offer is only one part of
conversion. Customers move through discovery, landing, registration,
verification, payment, initial use, support and repeat behaviour.
Friction at any stage can invalidate the apparent demand.
For market validation, I map the journey as a sequence of testable
events:
- the customer notices and understands the proposition;
- the source of the message is trusted;
- the landing experience matches the promise;
- registration works on the customer’s device and connection;
- required information feels proportionate and safe;
- identity checks can be completed;
- the preferred payment route is available;
- the transaction succeeds;
- the product delivers the promised outcome;
- support resolves predictable problems;
- the customer has a reason to return.
This approach produces better diagnoses.
If many customers click but few register, the issue may be message
alignment, page speed or form design. If registrations are strong but
funding is weak, payment access, trust or qualification may be the
constraint. If first use is healthy but retention collapses, the initial
incentive may be attracting the wrong behaviour.
A survey cannot locate these failures with the same precision.
Distribution must be
validated as a system
Businesses frequently validate a product and assume distribution will
follow.
That assumption is especially risky in emerging markets. The most
influential channels may be fragmented, relationship-led or difficult to
measure. A recognised media brand, local agent, comparison publisher,
creator or closed community may carry more trust than a technically
efficient advertising channel.
Distribution validation should test three things.
Access
Can the business reach partners or channels that serve the intended
audience? A contact database is not proof of access. The relationship
must be current, verified and commercially relevant.
Activation
Will those partners promote the product under acceptable terms?
Interest may disappear when the business introduces compliance
requirements, tracking rules, creative controls or payment
conditions.
Economics
Does the channel produce customers whose value exceeds commission,
media, incentives, fraud loss, support and operational cost?
In affiliate acquisition, I have seen the importance of connecting
partner activity to funded-customer behaviour. A tracked period with
3,189 registrations and 1,602 first-time depositors was meaningful
because it allowed the 50.2 percent registration-to-FTD conversion rate
to be examined alongside deposits and deal profitability. The
registration count alone would not have validated the channel.
The same logic applies outside iGaming. Distribution is validated
when the route to customer value works, not when a campaign produces
attention.
Trust needs its own
hypothesis
Trust is often treated as a branding outcome that can be addressed
after launch. In many digital markets, it is an operating condition.
Customers may need to trust that deposits will be recognised,
withdrawals will be processed, personal information will be protected,
prices will remain stable and support will respond. Partners may need
confidence that tracking is accurate and commissions will be paid.
Agents may need proof that the business will not bypass them after they
introduce customers.
Each trust concern should be written as a testable hypothesis.
Examples include:
- customers are more willing to register when a known local partner
explains the product; - transparent withdrawal conditions improve deposit completion;
- a visible local support route reduces abandonment;
- verified reviews increase conversion only when they answer specific
risk concerns; - partner activation rises when attribution and payment rules are
demonstrated before launch.
The test should connect the trust intervention to behaviour.
Awareness or positive sentiment may improve while transaction completion
remains unchanged.
Trust is validated through reduced hesitation, completed actions and
repeated use.
Use a validation scorecard
A validation project needs a decision framework before results
arrive.
Otherwise, teams reinterpret ambiguous evidence to protect the
preferred outcome.
A practical scorecard can include:
| Dimension | Evidence question | Example threshold |
|---|---|---|
| Problem | Does the target segment experience the problem frequently and urgently? |
Repeated observed behaviour, not stated inconvenience |
| Access | Can approved channels reach the segment? | Verified partners or campaigns with attributable reach |
| Conversion | Can customers complete the key journey? | Minimum rates for registration, verification and payment |
| Value | Does early customer behaviour support the model? | Defined repeat-use or revenue threshold |
| Economics | Does contribution remain positive after variable cost? | Agreed margin or payback boundary |
| Operations | Can the business deliver reliably? | Service, support and payment standards met |
| Compliance | Is the intended model permissible and controllable? | Legal review and control owners assigned |
| Learning | Is the evidence strong enough for the next decision? | Confidence level and unresolved risks documented |
Thresholds will vary by business. The discipline is to define them
before the pilot.
The scorecard should also support three outcomes, not two: proceed,
stop or redesign and retest. A failed assumption does not always
invalidate the market. It may show that the offer, payment route,
partner mix or operating model needs to change.
A pilot should answer a
narrow question
Many pilots are miniature launches with too many moving parts. When
results disappoint, no one knows why.
A better pilot isolates the most important uncertainty.
If payment fit is the concern, keep the proposition and acquisition
source stable while testing payment routes. If partner trust is the
concern, compare activation under different onboarding and reporting
conditions. If the market’s response to a price is uncertain, hold
fulfilment constant and test real purchase behaviour.
The pilot should define:
- the hypothesis;
- the target segment;
- the intervention;
- the expected behaviour;
- the success and stop thresholds;
- the measurement period;
- the known limitations;
- the decision that follows each possible result.
This creates learning that can survive beyond the pilot team.
Document what would
change the decision
Good validation is willing to be disproved.
Before a test begins, write down what evidence would cause the
business to delay, change or cancel the plan. This protects the process
from optimism after money and reputation have been committed.
It also improves executive conversations. Instead of debating whether
results “feel promising,” leaders can compare the evidence with agreed
decision boundaries.
The final report should separate:
- confirmed facts;
- customer or partner claims;
- observed behaviour;
- management interpretation;
- unresolved assumptions;
- risks that require specialist advice;
- the recommended next investment.
This distinction is particularly important when evidence comes from
several markets. A finding from Nigeria, for example, should not be
presented as proof of behaviour across Africa. It may create a
hypothesis for another market, but local validation remains
necessary.
Validation is a
capital-allocation discipline
The deepest purpose of market validation is not research. It is
better allocation of money, time and organisational attention.
A useful validation process tells leadership what has been learned,
how strong the evidence is and which uncertainty should be funded next.
It prevents a large launch from carrying risks that could have been
tested through a smaller commitment.
Surveys remain part of that process. They are useful when the
question concerns language, attitudes, awareness or self-reported
behaviour. They become dangerous when their results are asked to prove
demand, access or economics.
Market validation should move from opinion to commitment, from
commitment to behaviour and from behaviour to repeatable commercial
evidence.
The standard is not whether customers say the idea is good.
The standard is whether the proposed market, product and operating
model can create value together under real conditions.