Research before a launch answers three questions: is there demand for solving the problem, who is taking that customer today, and what will make your product different. Three separate pieces of work produce those answers — desk analysis of the market from open data, a competitor review, and demand testing with real people: interviews, a test page, prepayment.
Key takeaways
Research does not make a product successful; it changes decisions while they are still cheap to change. The most common root cause of company shutdowns has little to do with technology and everything to do with a product that missed a real market need.
The figures add up to more than one hundred per cent: in a review of 431 failed companies, several reasons were recorded for each. The practical conclusion for a launch is that interest in the idea is not the only thing to test — the economics have to add up on real numbers. How to calculate them is covered in our article on unit economics in plain terms.
The research itself closes four questions that otherwise get answered by guesswork:
Market size is described by three figures, and confusing them is expensive: teams plan against the largest and live off the smallest.
Bottom-up arithmetic is more reliable: how many businesses or people of your type exist in the country, what share of them fits the product's conditions, how much they pay for a similar solution today and how often that repeats. The top-down estimate — take a global report and multiply it by "one per cent of the market" — does not survive the first question about where that per cent came from.
Mistake Counting everyone who could theoretically buy as the market. The market is the people who have the problem, the money and the authority to decide. A head teacher may want your system while the district education authority pays for it: that is a different sales cycle and a different SOM.
For the Uzbek market, open sources are enough once you know where to look. The main rule is to record the source next to every figure: in a month you will not remember where it came from, yet the decision will rest on it.
Paid industry reports work as a reference point, but a decision cannot rest on them: the methodology is often undisclosed, and the country-level breakdown can be a calculated estimate rather than a measurement.
A competitor is any way the customer solves the problem today, including a spreadsheet, a messenger thread and the decision to do nothing. A list of three similar companies is not a competitor review; it is a list of similar companies.
The output of the review is one sentence answering the question "why will they choose us". Until that sentence exists, the launch waits: it is a question about the product, not about the advertising.
An interview tests the problem rather than the product: how the person runs into it, what they have already tried and what it cost them. Questions about the future — "would you buy this" — produce polite answers you cannot decide on.
Tip Compare two questions. "Would a system for tracking enquiries be useful to you?" — the answer costs nothing and commits to nothing. "Show me where last week's enquiries are kept and tell me how you worked through them" — the answer takes five minutes and shows the process as it really is.
Words are tested by deeds: an enquiry, a prepayment, a place on a waiting list. Testing through action costs less than development and produces a number rather than an impression.
The structure of such a page and the usual mistakes are covered in our article on landing pages for business. One caveat matters: the test measures demand for a promise. Only a product that keeps the promise confirms it.
The methods do not replace one another: desk analysis describes the market, interviews explain behaviour, action measures demand. The table shows what to expect from each method and what it will not show.
| Method | What it shows | What it takes | What it will not show |
|---|---|---|---|
| Desk analysis | Market size, structure and direction, barriers to entry | Open data, industry registers, an analyst's time | Whether anyone will buy from you specifically |
| Competitor review | Prices, positioning, weak points, sales channels | Websites and tariffs, reviews, a walk through their funnel | Their actual revenue and economics |
| In-depth interviews | How the problem arises and how it is solved today | Access to people with the problem and about an hour each | How widespread the described pattern is |
| Quantitative survey | How common the findings from interviews are | Hypotheses from interviews and a large enough sample | The reasons behind the answers people give |
| Test page with ads | Willingness to leave contacts and the cost of acquisition | A page, an advertising budget, analytics | Repeat purchases and retention |
| Search query analysis | The words people use for the problem and how often | Query statistics services and search suggestions | Demand where nobody searches in words |
The moment you collect contacts — a form on a test page, a survey questionnaire, a recorded conversation — you are processing personal data and you answer for it. The rules are the same for a live service and for a hypothesis test.
The law on personal data was amended by law ZRU-1125 of 26 March 2026. Under article 27-1, biometric and genetic data and data on users of telecom operators' services must be stored in Uzbekistan; other personal data may be stored abroad under the conditions listed in the same article — for example, when the country is recognised as providing adequate protection.
The output of research is a decision, not a report. There are three decisions, and each costs less than launching blind.
Record separately what convinced you: which quotes, which figures, which actions by real people. Six months later that list is the only defence against an argument that begins "but we agreed back then".
Syntra Systems does not sell market research as a separate service: the review of market and demand is part of the pre-project work on a product. Before designing a system we walk the same steps with the client — who the customer is, how they solve the problem today, what will make the offer different and whether the economics add up on their numbers.
From there the paths fork. If demand has to be tested before development, we build a test page with an enquiry form, analytics and Telegram notifications: a landing page starts from $1,200. If the business already runs and the question is where it loses money, a process review helps — it is covered in our article on the business audit.
Let’s discuss your project
Tell us what you need, and we will estimate the timeline and cost and suggest a solution.
Desk analysis shows the size and structure of a market but not the key question of whether anyone will buy from you. That answer comes from conversations with potential customers and from testing demand through action. A mismatch between product and market need is the most common root cause of company shutdowns.
Counting conversations is not the point; repetition is. While every new person still adds something unfamiliar, the segment has not been covered. Once the stories start repeating, there is enough to decide on that segment. For a different segment the count starts again.
You look at adjacent markets and at the current way the problem is solved: if people already pay for manual work, a contractor or someone else's service, demand exists and can be measured. A lack of statistics on a new category does not mean a lack of demand, but it raises the price of testing through action.
Research looks outward: the market, competitors, and whether people will pay for something that does not exist yet. A business audit looks inward at a working company and finds where money leaks in its processes. The first happens before a product launches, the second once sales are already running.
Yes, but a shorter version: the product hypothesis is already proven, and what needs testing is the transfer. Channels, payment methods, interface language, data requirements and the habitual way of solving the problem all change. Interviews in the new segment plus a demand test usually suffice.
Interviews are best run by the founder or the product team: what matters is not the conclusions in a report but the wording you hear first-hand. Desk analysis, data collection and scheduling are sensible to delegate. The launch decision stays with the company either way.
By action: a request from a test page, a waitlist signup, a prepayment. Prepayment is the strongest signal — someone pays before the product even exists, while an answer of ‘yes, I'd buy it’ settles nothing.
