How to Validate Demand for an IT Product: 6 Ways in Two Weeks

You can validate demand for an IT product before development in two weeks using six methods: problem interviews, analysis of search demand and competitors, a landing page with an offer, a clickable prototype, a manual service and presales. The first four methods show interest; the last two show willingness to pay. The “build or not” decision is made against pre-set thresholds, not on gut feeling after conversations.

This article is for founders, product managers and executives at IT companies in Kazakhstan who have come up with a new product or feature and want to understand whether it is worth spending the development budget on. We will go through each method with its criteria and give you a 14-day plan, a table for reading signals and common mistakes.

Why validate demand before development

Development is the most expensive way to find out that nobody needs your product. Even a small MVP requires several person-months of team work, while demand validation takes two weeks of the founder’s time and a small ad budget. We covered in detail what makes up the cost of development in our article on MVP costs in Kazakhstan.

Demand validation answers three questions:

1. Does a specific segment have a problem it is aware of and already solves somehow?
2. Is your offer different enough from the current solution for people to switch?
3. Are they willing to pay, and how much?

Before you start: the hypothesis and decision thresholds

The main mistake is starting validation without criteria and then interpreting any result as a success. So before the first interview, write your hypothesis in one line:

“[Segment] experiences [problem], currently solves it by [method], and is willing to pay [amount] for [outcome].”

Then set thresholds for each method: how many interviews must confirm the problem, what share of landing page visitors must leave an inquiry, how many prepayments or letters of intent you need to get. You choose the thresholds yourself based on the product’s economics. Other people’s “normal conversion rates” from English-language blogs have nothing to do with your segment and channel.

Six validation methods

1. Problem interviews

A conversation about the customer’s past experience, not about your idea. Ask how they solved the problem last time, how much time and money it took, what annoyed them and what they tried before. Do not show the product and do not ask “would you buy it”: people answer hypothetical questions politely, not honestly.

Criterion: interviewees name the problem on their own, without prompting, describe their current workaround and mention the money or time spent on it.

2. Analysis of search demand and competitors

See whether people search for a solution to the problem: query statistics in Yandex Wordstat and Google Keyword Planner, search suggestions, questions on forums and in specialized Telegram chats. Separately, list competitors and direct substitutes, including Excel, WhatsApp groups and contractors, and read reviews of them.

Criterion: there are queries phrased around the problem or the solution, and competitors show recurring complaints that your offer addresses. A complete absence of competitors more often means there is no market than that there is an open niche.

3. A landing page with an offer and an inquiry form

One page: who the product is for, what problem it solves, how it differs, roughly what it costs, and a “Submit a request” or “Sign up for early access” button. Drive traffic from one or two channels where your segment is. If you are unsure about the wording, make two versions of the offer and compare them on the same traffic source.

Criterion: the inquiry rate is at or above the pre-set threshold, and a noticeable share of those who left an inquiry agree to a follow-up call. Clicks without contact details are a weak signal.

4. A clickable prototype

A prototype of the key scenario in Figma or a similar tool. Give it to five to eight members of the segment with a task: “You need to do this.” Watch where they stumble and listen to what they say out loud.

Criterion: most people complete the scenario without hints and ask when they can try it on their own data. This method tests how clear the solution is, not willingness to pay, so it does not replace the fifth and sixth methods.

5. A manual service instead of a product

Provide the service the product promises manually: through spreadsheets, messengers and your own time. If the product matches customers with contractors, match them yourself. If it generates reports, compile them by hand.

Criterion: customers use the service repeatedly and are willing to pay at least a token amount. Bonus: you will see the customer’s actual process and understand which steps are really worth automating.

6. Presales

Offer to sell the product before it exists: a discounted prepayment with honest refund terms, a letter of intent or, for B2B, a pilot agreement with payment after launch.

Criterion: a pre-set number of payments, signed letters or agreements. This is the strongest of the six signals: words cost nothing, while money and signatures do.

Summary by method:

Method	What it tests	Signal strength	Time
Problem interviews	Whether the problem exists and how acute it is	Medium	5–7 days
Search and competitor analysis	Problem awareness, market size	Medium	1–2 days
Landing page with an offer	Interest in a specific proposal	Medium	5–7 days of traffic
Clickable prototype	Clarity of the solution	Low for demand	2–3 days
Manual service	Value of the outcome	High	5–10 days
Presales	Willingness to pay	Highest	In parallel with the others

A 14-day validation plan

Days	What we do	Outcome
1	Formulate the hypothesis, segment and decision thresholds	One page with criteria
2	Analyze search demand, competitors and reviews	A list of substitutes and complaints
2–3	Recruit interviewees, prepare the interview script	15–20 scheduled calls
3–7	Conduct interviews, writing down takeaways after each one	A problem map and customer quotes
6–7	Build a landing page in the customers’ language from the interviews	A landing page with a form and analytics
8–13	Drive traffic to the landing page, call those who left inquiries	Conversion and qualified leads
8–10	Build a prototype and test it with 5–8 people	A list of clarity issues
9–13	Launch a manual service and presales for the most interested	Payments, letters, repeat requests
14	Compare results with the thresholds and make a decision	“Build,” “change the hypothesis” or “stop”

The numbers in the plan are a guideline for one segment and one person running the validation. If the segment is B2B and decisions take a long time, extend presales into the next cycle, but have the first conversations about money within these same two weeks.

Kazakhstan context. The market is small, and in B2B, decisions are often made through acquaintances and referrals. So it is easier to reach interviewees through personal connections and industry communities than through cold emails, but you need to track separately for sample bias toward “your own” people. Many interviewees will prefer messages and voice notes on WhatsApp to a video call, and that is fine. For a mass audience, test the landing page in Russian and Kazakh: the language of the offer affects the response. The inquiry form collects personal data. If it matters to store the inquiry database in Kazakhstan, take this into account when choosing a landing page builder and CRM.

How to read the signals

Signal	What it means	Decision
People name the problem themselves, have a workaround and have spent money	The problem is real and recognized	Move on to the offer and presales
People acknowledge the problem only after a prompt	The problem is not a priority	Look for another segment or another problem
Many clicks on the landing page, few inquiries with contacts	The headline is interesting, but the offer is not	Rewrite the offer, test the price
There are inquiries, but people do not agree to a call	Weak motivation or the wrong segment	Refine the segment and traffic channel
People complete the prototype but are not willing to pay	The solution is clear, but the value is small	Reconsider the outcome the product promises
People use the manual service repeatedly	Value is confirmed	Automate the most labor-intensive steps
There are prepayments or signed letters	Willingness to pay is confirmed	Plan an MVP for this segment
Signals from different segments contradict each other	Audiences are mixed	Choose one segment and repeat the cycle

How to make the decision on day fourteen

The decision is made in a single meeting, against the threshold table you wrote down on day one. There are three options.

Build. The thresholds are met on at least one strong signal: there are prepayments, signed letters or repeat use of the manual service, backed up by interviews. The next step is to define an MVP only for the scenario people have already paid for and launch it with the people on your inquiry list. Everything that was not validated goes to the backlog.

Change the hypothesis. The problem was confirmed, but the offer, the price or the segment did not work. This is a normal and common outcome. Rework one variable, not all of them at once: a different segment with the same offer, or a different offer for the same segment. Then run another short cycle, this time without repeating the market analysis.

Stop. People do not name the problem without prompting, there are few inquiries, and nobody is willing to pay. This is also a result: you spent two weeks instead of several months of development. Keep the interview notes and the contact list: they will come in handy for your next idea in the same industry.

To keep the decision from turning into a debate about taste, agree in advance who makes the final call, and record the outcome in writing along with the numbers it is based on.

Common mistakes

• Testing on friends and colleagues. They want to support you and answer accordingly.
• Asking “would you buy it.” Hypothetical answers are almost always positive and mean almost nothing.
• Setting thresholds after the results. Then any result can be called a success.
• Mixing segments. Answers from small businesses and corporations in one sample produce an average that describes no one.
• Stopping at interest. Likes, subscriptions and clicks are not money. Without presales or a manual service, willingness to pay has not been tested.
• Building the landing page in your language rather than the customer’s. It is better to take the wording for the offer verbatim from the interviews.
• Not recording data. Inquiries and agreements should live in a CRM or at least a spreadsheet, not in personal chats.

If validation succeeds and demand is confirmed, the next step is to choose the channels through which the product will get customers. That is the subject of our article on GTM for SaaS in Kazakhstan.

How we do it

We start with a hypothesis diagnostic: together with the team, we formulate the segment, the offer and the decision thresholds, then run a two-week validation cycle on one segment. If the signals are strong, we build an MVP only for the confirmed scenario and launch it with those who have already left an inquiry or made a prepayment. We scale what the data has confirmed, not what we originally wanted to build.

If you want to go through demand validation and a product launch with us, start with the MVP development and product launch page: it describes the process and what you get at each stage.

---

Sources. The timelines and thresholds in this article are Gotomarket’s working model, not market statistics.

Terms in plain words

MVP (Minimum Viable Product)

The minimal working version of a product for testing an idea with real users. Demand validation is there so you do not spend your budget on an MVP nobody needs.

Landing page

A single-page website describing the offer, with an inquiry form. At the validation stage, it shows whether people leave their contact details even though the product does not exist yet.

Offer

The way the proposal is phrased: who the product is for, what problem it solves, how it differs and roughly what it costs. It is best written in the customers' own words from the interviews.

Segment

A group of customers with a similar problem and a similar way of buying. If you mix segments, validation results average out and describe no one.

Conversion

The share of visitors who took the desired action, for example left an inquiry on the landing page. The conversion threshold is set in advance based on your own economics.

Clickable prototype

An interface mockup through which you can 'walk' a scenario by clicking buttons, even though there is no working code behind it. It tests clarity, not willingness to pay.

Concierge service (manual MVP)

You manually provide the customer with the service the product promises, using spreadsheets and messengers. This tests the value of the outcome before development.

Letter of intent

A document in which a client company confirms its intention to buy the product after launch. For B2B, it is a strong demand signal, almost as good as prepayment.