Customer Development: The Four Stages and How to Use Them

Customer Development is a method for testing a startup’s business-model assumptions with customers before treating those assumptions as facts. It runs alongside product development; it does not replace building the product.

Steve Blank’s model has four stages: Customer Discovery, Customer Validation, Customer Creation, and Company Building. In his original Customer Development explanation, Blank emphasizes that the model is iterative rather than a one-way sequence.

That point matters. A startup does not “finish interviews” and graduate permanently to growth. New segments, products, prices, and channels can send the team back into discovery and validation.

The four Customer Development stages

These stage descriptions follow Blank’s model, but the exact experiments vary by business. Enterprise software, a local service, a marketplace, and a regulated product should not use identical scripts or evidence standards.

Stage 1: Customer Discovery

Customer Discovery turns a founder’s beliefs into explicit hypotheses and tests them outside the company.

Write the hypotheses

Cover at least:

• customer and user;
• painful job or problem;
• current alternative;
• trigger and urgency;
• proposed value;
• channel and buying process;
• revenue and cost logic;
• technical or delivery feasibility; and
• major legal, regulatory, or adoption constraints.

Do not make the list so large that it becomes a business-plan exercise. Rank the assumptions by consequence and uncertainty.

Investigate behavior

Ask customers about a recent real event:

• What triggered the problem?
• What did they do first?
• Which people became involved?
• What did the workaround cost in time, money, delay, or risk?
• Which alternatives did they consider or reject?
• What budget, approval, or procurement path applied?

The customer discovery questions guide provides a deeper interview structure designed to avoid compliments and hypotheticals.

Test a small solution

Use the lowest-cost experiment that can reveal the next important behavior: a manual service, prototype, sample, workflow walkthrough, landing page, paid diagnostic, or pre-sale.

Customer Discovery does not require a fully built product. It does require enough specificity for the customer to react to the actual tradeoff.

Discovery exit evidence

Do not use a universal interview count. Move forward when the team can describe a narrow customer, recurring problem, current behavior, buying context, and testable solution with enough consistency to justify a validation offer.

Open contradictions should be visible, not averaged away.

Stage 2: Customer Validation

Customer Validation asks whether the startup can create a repeatable and scalable sales model. It is stronger than proving that some people have the problem.

Validation evidence may include:

• customers from the intended segment commit time, data, reputation, or money;
• the team can identify the buyer and approval path;
• similar problems lead to similar offers;
• the product or service can be delivered successfully;
• customers reach the intended value and continue or expand; and
• acquisition and delivery economics are becoming understandable.

One founder-network sale may be meaningful, but it does not prove a repeatable channel. A paid pilot may validate willingness to test without validating renewal. Keep each claim as narrow as the evidence.

Use the customer validation guide for an evidence ladder and the paid-pilot examples for bounded B2B tests.

Validation exit evidence

The team should be able to explain:

• who repeatedly buys;
• why they buy now;
• how the opportunity is sourced and advanced;
• what must happen for delivery to succeed;
• which behaviors indicate value; and
• which economics and constraints still require testing.

“Repeatable” does not mean automatic or risk-free. It means the team has a model it can describe, measure, and test with people beyond the original friendly network.

Stage 3: Customer Creation

Customer Creation develops demand for the validated offer. The method depends on the market type: entering an existing category, creating a new category, resegmenting a market, or serving a niche each changes positioning and adoption.

Key work includes:

• choosing the initial segment and message;
• matching channels to how buyers actually discover and evaluate solutions;
• building credible proof;
• connecting acquisition to activation and retention;
• measuring cost and quality, not only lead volume; and
• increasing spend in stages as evidence holds.

This is where founders often scale too early. A campaign can create many leads and still hide a weak product, unclear buyer, or unsustainable sales process.

The startup go-to-market guide helps translate validated learning into a focused market motion.

Stage 4: Company Building

Company Building shifts the organization from primarily searching for a model toward executing and improving one. The startup adds functions, management systems, and specialists because recurring work and scale require them—not because a generic org chart says it is time.

Practical signs include:

• functional priorities can be derived from one company strategy;
• demand and delivery are forecastable enough to plan capacity;
• customer and financial definitions are stable;
• quality and risk controls need dedicated ownership;
• decisions can move beyond the founders without losing context; and
• new hires enter documented, repeatable workflows.

The search never disappears. Mature operating systems should preserve customer learning rather than isolate it in a research team.

How to run Customer Development week by week

Maintain an assumption board

For each important belief, record:

Separate notes from conclusions

Store what the customer did and said before summarizing the lesson. Quote carefully, preserve context, and distinguish one observation from a pattern.

Review contradictions

The outlier may reveal a different segment, buyer, workflow, or risk. Do not force every interview into the same story.

Update the business model

Customer Development should change decisions. After each evidence cycle, update the relevant customer, value proposition, channel, pricing, cost, or delivery hypothesis.

Blank later described Customer Development as a way to test the assumptions behind a business model and turn new evidence into pivots; see his Business Model and Customer Development stack.

Common Customer Development mistakes

Treating the framework as linear

The four stages show different jobs, not irreversible graduation. Return to discovery when a core assumption changes.

Outsourcing all customer contact

Early on, founders need direct exposure to the words, workflow, emotion, and constraints behind the data. A research agency or salesperson can support the process but should not become a filter that removes learning.

Asking for opinions about the idea

Future intent is weak evidence. Investigate past behavior and design a real next step.

Counting interest as validation

Email signups, survey enthusiasm, and social engagement can justify another test. They rarely prove a repeatable business by themselves.

Scaling demand before delivery works

Growth spending magnifies both value and defects. Validate onboarding, delivery, and retention before treating lead volume as success.

Using one evidence standard for every business

A medical device, consumer app, consulting offer, and marketplace have different technical, regulatory, sales, and usage risks. Define evidence for the actual model.

Customer Development as a decision system

The framework is valuable because it gives uncertainty a workflow:

Hypothesis → field evidence → experiment → decision → updated model

The output is not a pile of interview notes. It is a better decision about who to serve, what to offer, how to reach them, and whether the business has earned the next investment.