Friends find your idea interesting. A few people say they might buy. You have identified a sizeable market. These are encouraging signals, but they do not yet answer the essential question: will enough customers you can actually reach pay a price that sustains your business?
Useful market research is not measured by the thickness of a report. It should help you decide what to sell, to whom, at what price and through which channel, while identifying assumptions that still need testing.
At a glance
- Start with a specific decision: choosing a customer segment, location, price or offer.
- Combine existing data, customer interviews and real-world tests.
- Study the solutions customers already use, including buying nothing.
- Distinguish positive feedback, stated intentions and concrete commitments.
- Connect estimated demand to margins, acquisition costs and delivery capacity.
1. Define the assumptions you need to test
An idea becomes testable when it identifies a customer, problem and proposition. “Helping SMEs become better organised” is too broad. “Providing small maintenance companies with a monthly service to prepare administrative paperwork for jobs” gives you specific people and situations to investigate.
Write your main assumptions as statements:
- managers spend several hours a week on this task;
- the problem causes measurable delays or losses;
- the person experiencing it can authorise the purchase or influence the decision-maker;
- they are willing to pay your proposed price;
- you can reach them without spending more than the expected margin.
Rank these assumptions by importance and uncertainty. If the business depends on a licence, scarce premises or a single supplier, check that constraint before conducting a lengthy customer survey.
Also define what would make you change your mind. Without decision criteria, it is easy to focus only on answers that support the original idea.
2. Define the market you can actually reach
A national figure provides context, but is not automatically your initial market. A local business depends on travel distance, opening hours and a particular clientele. A business-to-business service depends on customer size, organisation and purchasing processes.
For an initial estimate, distinguish:
- everyone with the relevant need;
- those your offer could realistically suit;
- those you can reach with your current resources;
- those you can serve without compromising quality.
Federal Statistical Office data, particularly structural business statistics, along with cantonal and municipal statistics, can help define a customer base. Check the date, geographical scope and unit: an establishment is not always a company, and a job is not a resident.
Example: a market of 2,000 businesses is not a sales forecast
Suppose an area has 2,000 establishments in sectors close to your target. After filtering, 250 match your intended customer size and service. You have a realistic way to contact 80 during the first few months.
Your sales plan should explain how those 80 contacts might become meetings and then customers. Writing “we will capture 1% of the market” explains neither the resources required nor the time needed.
3. Study your customers’ alternatives
Competitor analysis extends beyond companies selling exactly the same thing. Your prospective customer might do the work themselves, ask an employee, use software, buy part of a service or simply tolerate the problem.
Build a comparison from observable evidence: websites, public offers, visits, conversations and accessible customer feedback. Do not present assumptions about a competitor’s margins or sales as facts.
| Alternative | What the customer values | Possible limitation to check | Implication for your offer |
|---|---|---|---|
| Doing it themselves | No additional bill | Time required and mistakes | Quantify the time genuinely saved |
| Established provider | Trust and continuity | Overly broad service or limited availability | Define a focused, reassuring service |
| Digital tool | Autonomy and predictable pricing | Setup and learning curve | Show the value you add beyond the tool |
| Changing nothing | No transition effort | A tolerated problem with ongoing costs | Check whether the need is a sufficient priority |
Online reviews can reveal frustrations, but may not represent the whole customer base. Use them to develop questions, rather than assuming all customers share the same view.
4. Conduct customer interviews that teach you something
Speak to people who match your target, beyond friends and family. An initial series of interviews can reveal recurring problems without claiming to be statistically representative.
Start with past behaviour:
- “When did you last experience this problem?”
- “How did you solve it?”
- “How much time or money did it cost?”
- “Who chose the solution?”
- “What would stop you changing your approach?”
These questions generally tell you more than “Do you like my idea?” Present your solution after understanding the current situation so it does not steer the entire conversation.
After each interview, keep a short record: respondent profile, recent problem, current solution, observed cost, decision-maker and next point to verify. Separate customer quotations from your interpretation. Distinguish recurring themes from isolated cases. An enthusiastic but highly specific request can draw you into a bespoke offer that is difficult to sell to anyone else.
When should you use a survey?
A questionnaire becomes useful once you know what you want to measure: purchasing frequency, selection criteria or the distribution of different uses. Ambiguous questions produce apparently precise numbers with little practical value.
Explain how respondents were recruited. Fifty answers from an enthusiast group do not necessarily describe a canton’s population. Present results with that limitation and avoid treating a convenience sample as a universal finding.
5. Test a real offer and price
The test should require an effort proportionate to the commitment you want to measure: booking a meeting, requesting a detailed quotation, accepting a paid pilot or placing an order you can fulfil.
A landing page can test a message. By itself, it does not prove that customers will buy or continue using the service. Be transparent about what already exists, delivery times and conditions. Do not imply availability or testimonials that do not exist.
To set your price, present a sufficiently specific offer: deliverables, limits, timing, terms and amount. “Would you pay between CHF 100 and CHF 1,000?” tells you little. “Here is what is included for CHF 390 a month” creates a concrete discussion.
Measure the whole customer journey
Consider this fictional business-to-business service test:
| Stage | Observed result |
|---|---|
| Businesses contacted | 40 |
| Interviews completed | 12 |
| Proposals sent | 6 |
| Paid pilots accepted | 3 |
The result shows that three customers accepted an offer under the test conditions. It does not establish that 7.5% of all businesses in Switzerland will buy. Targeting, personal relationships and sales effort may have strongly influenced the outcome.
Also examine why others declined: no need, poor timing, price, lack of trust or an unsuitable offer. These reasons require different responses.
6. Connect demand to profitability
Finding customers is not enough if each sale consumes too much time or too many resources. Your test must measure delivery costs, not just acceptance of the price.
Suppose a subscription sells for CHF 390 per month excluding VAT. You estimate five hours of work at an economic cost of CHF 40 an hour, plus CHF 30 for tools and variable expenses. Variable cost is CHF 230, leaving a contribution margin of CHF 160 per customer per month to cover fixed costs and then generate profit.
With CHF 2,400 of monthly fixed costs not already included above, you need 15 customers to cover them: CHF 2,400 ÷ CHF 160 = 15. Those customers already represent 75 hours of monthly delivery work, before sales, administration and unexpected tasks.
If delivery actually takes seven hours, variable cost rises to CHF 310, the margin falls to CHF 80 and you need 30 customers. That requires 210 hours of delivery work per month. The same market-accepted price can therefore produce a viable or unworkable model depending on actual delivery time.
Add customer acquisition cost: the spending required to win a new customer. If acquisition costs CHF 480 and monthly contribution margin is CHF 160, three months of margin are needed to recover that cost before other factors. If customers leave after two months, the model has a problem even while revenue is growing.
These examples are reasoning tools, not pricing or productivity benchmarks. Adapt them to your activity and avoid counting the same costs twice.
7. Make a decision and document what remains uncertain
Finish with a short synthesis separating observed facts, assumptions and decisions.
| Question | Evidence needed | Possible decision |
|---|---|---|
| Is the problem a priority? | Recent situations, costs and existing solutions | Keep or change the target customer |
| Is the price acceptable? | Accepted quotations or comparable paid tests | Adjust the offer, price or segment |
| Can you find customers? | A tested channel and measured sales effort | Choose a channel or review acquisition cost |
| Can you deliver at an adequate margin? | Actual time and costs | Simplify the service or reorganise delivery |
| Can you finance the launch? | A budget and cash collection timetable | Launch, narrow the scope or postpone |
Mixed results do not mean the research failed. They can prevent an expensive lease, an overly broad offer or premature investment. A better-founded decision is the useful outcome.
From market research to a measured launch
Market research does not eliminate uncertainty. It reduces unnecessary bets and helps identify what you still need to learn. Start with a specific customer segment, test your offer against actual behaviour and connect every sale to its costs.
Once these elements are sufficiently clear, build your cash flow forecast and include the findings in your business plan. Entreprendre.ch helps you prepare and set up your business.
Frequently asked questions
Does a small business need market research?
Yes, but it does not necessarily need an expensive study. Targeted desk research, interviews and a practical test can already inform an important decision.
How many people should I interview?
There is no universal number. Qualitative interviews seek to understand behaviour. Statistical estimates require suitable sampling and methodology. Do not present a few interviews as representative of the entire market.
Is a positive survey enough to launch?
No. Stated intention is not a purchase. Test the price, conditions, access to customers and ability to deliver at a sufficient margin.
Where can I find Swiss market data?
The Federal Statistical Office, cantonal and municipal statistics offices and industry bodies can provide useful data. Check the date, scope and definitions of indicators.
What if the results are disappointing?
Identify the assumption that does not hold: need, target customer, price, channel or production cost. Where possible, change one element at a time and test again before committing more resources.
Sources and resources
- SECO SME Portal: market analysis, quantitative and qualitative approaches.
- Federal Statistical Office: establishments and jobs by municipality and economic sector (in French) (in French).
- SME Portal: preparing a marketing strategy (in French).
Resources checked in September 2026. All commercial and financial examples are fictional.
Romain Prieur is a Swiss certified public accountant and the founder of Entreprendre.ch. With more than ten years of experience in auditing and supporting Swiss businesses, he advises entrepreneurs on company formation, accounting and taxation.
