How to Validate a Startup Idea Before Development

Validating a startup idea is a series of low-cost tests before any code is written. You form hypotheses, talk to your audience, assess the market and launch a simple MVP. This article covers the methods, examples, decision criteria and a checklist.

Why You Should Validate a Startup Idea Before Development

The main reason companies shut down is a lack of market need. According to CB Insights, about 35-42% of startups close for this reason. The exact share depends on the version of the report. Founders spend years building a product nobody needs.

Development takes money and time. Why validate a startup idea in advance? To avoid spending time on a product with no demand. Such validation lowers risk and exposes the weak spots of the idea. What the team learns before writing code:

• how acute the pain is for the target audience;
• whether buyers are ready to pay or not;
• how much money is needed before the first sales;
• whether there is real interest in the market;
• which business risks to close first.

Tests take weeks, while development takes months. The first sales and pre-orders confirm demand better than any survey. This approach saves budget and keeps founders disciplined. The founder makes decisions based on facts.

What Customer Development and an MVP Are

Customer Development is a methodology by Steve Blank. It is built around talking to customers before a product is created. The author described it in the book "The Four Steps to the Epiphany" in 2005. The main principle: assumptions are tested on real people.

The methodology has four stages: discovering needs, validating the solution, creating demand and building the company.

An MVP is a minimum viable product. The term was popularized by Eric Ries in the book The Lean Startup in 2011. A trial product contains the minimum set of features. It tests the key hypothesis together with real users.

CustDev answers the question of what people need. A minimal version shows whether they will buy an answer to their task. Both methods provide data for the next step.

How to Test a Hypothesis Quickly

Fast validation runs in four stages. First, the team formulates assumptions. Then it runs conversations and demand tests. Each stage provides data for a decision. Start with the riskiest assumption. If it is not confirmed, the other checks lose their meaning.

How to Formulate Hypotheses

A hypothesis is a testable statement about the audience, the pain or the product. A convenient format: "We believe that a segment experiences a difficulty and is willing to pay for such a product." To formulate hypotheses, identify four types:

1. Problem hypothesis: the problem exists and matters.
2. Solution hypothesis: your product solves this problem.
3. Value hypothesis: the buyer sees the value proposition and is willing to pay.
4. Channel hypothesis: people can be reached at a reasonable price.

Set a success criterion for each statement. For example, seven out of ten respondents confirm the pain. Test assumptions about business models last.

Problem Interviews With the Target Audience

A problem interview finds out how people handle a task today. In-depth interviews give more data than questionnaires. Interviews should be held with representatives of the target audience.

You need to understand whether the problem exists. Ask: "Tell me how you solved this task the last time." Find out how people solve it now and how much they spend. Do not mention your product until the end of the conversation.

Examples of follow-up questions: "What did you do next?", "How long did it take?", "What was the most inconvenient part?" Such questions reveal real behavior.

The conversation consists of five blocks:

• introduction and context;
• a description of the last situation;
• current methods and costs;
• the result and dissatisfaction;
• a request for contacts of other respondents.

A good benchmark is 10-15 conversations. After that, the answers start to repeat. Record the conversations and group the answers by topic.

Common Mistakes in Interviews

Many founders run a conversation like a presentation. This distorts the data. The most common mistakes:

• leading questions like "Would it be convenient for you?";
• questions about future behavior;
• talking about your own product at the start of the conversation;
• surveying friends and relatives;
• no recording and no analysis of the answers.

Questions about past behavior are more reliable than opinions about the future. A person cannot honestly answer the question "Would you buy it?" Ask about real actions and spending.

Quick Demand Tests Without Code

A demand test shows real interest without writing code. It helps you quickly test demand and costs little. The main formats:

• a landing page with a description and an application button;
• a small ad campaign sending traffic to the site;
• a pre-order or a waiting list;
• a post in niche communities;
• a clickable prototype in Figma;
• delivering the service manually to the first buyers.

The site is built on a constructor, for example Tilda or Webflow. The main metric is the share of site visitors who leave a contact or a payment. Set a target value before launch. Record the traffic source and the cost per application, so you can see which channel works.

Example: a lunch delivery service for offices is tested with a page showing a menu and an order form. The first applications arrive faster than a finished app. If there are no applications, you avoid the cost of code.

How to Assess the Market, Demand and Competitors

Such an assessment shows whether there are enough buyers for the business. Competitor analysis confirms that demand exists. These data are needed by investors and founders.

Market Size and Demand

Market size is assessed with three indicators. TAM is the total sales volume of the category. SAM is the part of that volume available to your product. SOM is the share you can capture in the coming years.

Calculation example: a city has 10,000 offices, and 3,000 of them fit your size criteria. In the first year you can realistically connect 100. Then TAM equals 10,000, SAM equals 3,000, and SOM equals 100 offices.

Demand is checked through search queries. Yandex Wordstat and Google Trends work for this. Industry reports and open statistics provide additional data. Compare the estimate with the respondents' answers. A discrepancy points to an error in the calculations. A top-down estimate is built from reports, and a bottom-up estimate from the number of buyers and the average check.

Direct and Indirect Competitors

Direct competitors meet the same need for the same people. Indirect competitors do it in another way. For example, for a note-taking service, an indirect competitor is a paper notebook.

Build a table with six columns: name, price, features, strengths, weaknesses, positioning. Separately write down each player's value proposition. If there are no competitors, check the demand again.

To find competitors, use search results, app catalogs, reviews and niche forums. Study what people praise and criticize in each service. These data show unmet needs.

SWOT Analysis of the Idea

A SWOT analysis helps assess the idea across four groups of factors:

• strengths: unique competencies and resources;
• weaknesses: a lack of experience, money or time;
• opportunities: market growth, new technologies;
• threats: competitors, regulation, changes in demand.

Strengths and weaknesses relate to the project itself. Opportunities and threats relate to the external environment. Fill in the table using the data from conversations and competitor analysis. For example, a weakness is having no developer on the team.

Express Scoring: How to Assess an Idea Quantitatively

Scoring turns impressions into numbers. It helps compare several business ideas by the same criteria. The method below serves as a working benchmark.

Assessment Criteria: Sales Market, Capital, Resources, Founder Commitment

Assess the project against four criteria. Give each a score from 1 to 5:

• sales market: size, growth and accessibility of customers;
• capital: whether the money lasts until the first sales;
• resources: competencies, a team and access to the audience;
• founder commitment: the time and effort they put into the project.

It is better to score together with a partner or an expert. This reduces the influence of personal expectations. Write down the reasoning for each score.

How to Calculate the Result

Add up the scores for the four criteria. The maximum result is 20. Make the decision using the table.

Total score	Decision
16-20	The project is strong, move on to demand tests and a minimal version
11-15	Strengthen the weak criteria and repeat the assessment
4-10	Reconsider the concept or the customer segment

A criterion scored 1 blocks the launch even with a high total. For example, there is no capital, or the founders work only in the evenings. Scoring complements conversations and demand tests.

Example of an assessment: sales market 4, capital 3, resources 4, commitment 5. The total is 16, so the project counts as strong. The weak criterion is capital, so settle the question of funding first.

MVP Types With Examples

The format of a minimum viable product depends on the product and the budget. Five common formats are described below. Each comes with an example from the practice of well-known companies. The table compares the formats.

Format	What it tests	Cost
Landing page	Interest in the description	Low
Wizard of Oz	Willingness to pay for the result	Medium
Concierge	The value of the service	Low, but a lot of manual work
Piecemeal	Whether a combination of services works	Low
Single-Feature	The main feature	Medium

The decision about the format depends on the hypothesis. To test interest, a landing page is enough. To assess willingness to pay, choose manual formats.

Landing Page and MVP Content

A landing page describes the product and collects applications. At an early stage, Dropbox showed its idea in a demo video. According to founder Drew Houston, the waiting list grew from 5,000 to 75,000 people overnight. In 2010, Buffer tested its concept with two pages: a description and pricing.

Wizard of Oz

The user sees a finished product, while the processes are carried out manually. Zappos started in 1999 with photos of shoes from stores. Founder Nick Swinmurn bought a pair in a store after an order came in. This is how he tested people's willingness to buy shoes online.

Concierge MVP

The founder serves people personally and by hand. This shows what value people are willing to pay for. An example is Food on the Table. The founder manually composed menus and shopping lists for the first customers.

Piecemeal MVP

The product is assembled from ready-made services: forms, spreadsheets, payment systems. Code is kept to a minimum. Groupon started as a WordPress blog. Coupons were sent out manually as PDF files.

Single-Feature MVP

The product implements one feature. Instagram started as a project called Burbn with many capabilities. The founders kept photo posting and filters. This focus simplified the work and sped up the launch.

Where to Start Developing a Startup

After the assumptions are validated, work on the product begins. Startup development starts with four steps. Each step relies on test data.

Define the Target Audience and the Problem

Describe the customer portrait: segment, tasks and current methods. Formulate one problem that the product removes. Use the conversation data. This wording becomes the basis for the value of the product.

List the Features and Define the MVP Scope

Write out all the product features. Divide them into must-have, nice-to-have and postponed. Include only the must-have ones in the first version. The scope of the starting set defines the timeline and the budget.

Choose a Development Method and a Team

For a minimal version, the flexible Agile method works well. It provides short cycles and fast feedback. The team is assembled by roles: developer, designer, tester and manager. Hiring options: in-house staff, freelancers or a studio. Without technical expertise, a turnkey studio is a good choice.

Run Alpha and Beta Testing

Alpha testing takes place inside the team. Testers look for bugs before the product reaches users. Beta testing is run on a limited group of users. Collect feedback and usage metrics. After the beta test, compare the metrics with the thresholds and make the launch decision.

Metrics and Criteria: Continue, Pivot or Stop

Decisions are made based on metrics set in advance. Thresholds are set before the test begins. The main indicators:

• landing page conversion to applications;
• the share of respondents who confirmed the pain;
• the share of pre-orders and payments;
• customer acquisition cost;
• user retention after a week and a month.

The main question is whether people pay for the solution. The decision to scale is made after repeatable sales. Results are compared with the thresholds using the table.

Test result	Action
Metrics reached the thresholds	Continue and expand the test
The pain is confirmed, the product is rejected	Change the product format (pivot)
The pain is not confirmed	Stop or change the segment

A pivot is a change of hypothesis while keeping the goal. If the metrics grow, you can move on. Stopping at an early stage saves strength for a new business.

Example of thresholds: landing page conversion of at least 5%, three pre-orders out of ten conversations. Specific values depend on the industry and the price. Record them in writing before launch.

Common Mistakes When Validating an Idea

Mistakes at this stage lead to unnecessary expenses. Six mistakes occur most often:

1. Testing only on friends and acquaintances.
2. Looking for confirmation instead of facts.
3. Starting to code before the first conversations.
4. Ignoring direct and indirect competitors.
5. A starting set of features that is too large.
6. No success criteria before the test begins.

Many founders write code instead of talking to people. The result is an expensive product with no demand. Set metrics and thresholds in advance.

Startup Idea Validation Checklist

Go through the list before starting work. Every item needs an answer:

• assumptions are written down and have success criteria;
• 10-15 conversations with respondents have been held;
• the pain is confirmed by the target group;
• the market size is estimated using TAM, SAM and SOM;
• competitors are studied;
• a SWOT analysis and scoring are done;
• the demand test produced applications or pre-orders;
• the starting set is limited to must-have features;
• metrics and thresholds are defined.

If an item has no answer, return to the relevant section. Then repeat the assessment.

Frequently Asked Questions

How much does it cost to validate a startup idea?

The cost depends on the format of the tests. Conversations and manual checks are almost free. The main expenses are advertising, a website builder and the team's time.

How many conversations should you hold?

A good benchmark is 10-15 conversations with representatives of the target audience. Continue until the answers start to repeat.

Can you validate an idea without code?

Yes. A landing page, a pre-order, delivering the service manually and a clickable prototype all work.

How does an MVP differ from a prototype?

A prototype shows the interface and the logic of the screens. An MVP works and brings value to real users.

Is a technical co-founder needed to validate an idea?

Most tests need no code. A technical specialist is needed at the stage of building the product.

When is it time to start development?

When the pain is confirmed, people are willing to pay, and the test metrics have reached the thresholds.

Can AI be used when validating an idea?

Yes. Neural networks speed up the analysis of conversation transcripts and the search for competitors. The decision is made by a person based on data from real customers.

What is a pivot?

A pivot is a change of hypothesis while keeping the company's goal. For example, the team changes the segment or the product format based on test results.

How does CustDev differ from market research?

CustDev tests hypotheses in a live conversation with people. Market research collects statistics from large samples. The methods complement each other.

Conclusion

The work starts with assumptions, conversations and demand tests. Then you assess demand, calculate the score and choose the format of the minimal version. The final decision is made by the metrics. Defence.Investments builds an MVP turnkey and helps launch a startup turnkey.

Get a free project assessment