Software Growth

Customer development

Customer development is Steve Blank's method for testing a startup's business model by talking to customers before and while you build.

Customer development is a process for finding out whether anyone wants what you plan to build, before you spend a year building it. Steve Blank developed it and set it out in his 2005 book The Four Steps to the Epiphany. His core line is that there are no facts inside the building. Your assumptions about customers, pricing and demand are guesses until you test them outside with real people.

The four steps

  1. Customer discovery. Turn your hypotheses about the problem, customer and business model into facts. You interview people and test whether the problem is real and whether your solution fits.
  2. Customer validation. Learn to sell to early customers and repeat the sale. The point is proof that the model works, not revenue itself.
  3. Customer creation. Start spending on marketing to build demand.
  4. Company building. Move from a founding team to a company with functions and processes.

The steps loop, and going back is normal. Blank pairs the process with the business model canvas, and his work is the base for the lean startup.

How to run customer discovery

Write down your hypotheses: who has the problem, how they handle it today, what they would pay. Then get out and test them. One approach in Blank's material is to hold roughly 10 to 15 customer conversations a week. Record what you learn and change the plan when facts disagree with it. A failed hypothesis is a result, not a setback.

The Mom Test

Customer development depends on getting honest answers, and people tend to be polite. Rob Fitzpatrick's book The Mom Test is a short guide to asking questions that produce real information even when the person is trying to be nice. The principles are practical: ask about their life and past behavior, not about your idea, avoid biased questions, and look for signs of real pain and willingness to pay. The book is used in teaching at places like Harvard and UCL.

Example

You plan a tool to automate client reports for freelance designers. You interview 15 designers. Twelve say reporting is annoying, which sounds promising. But only 3 have tried a tool, and 1 pays for one. 3 divided by 15 is 20 percent, and 1 divided by 15 is about 7 percent. The pain is mild, and few will pay. Better to learn that in two weeks than after six months of code.

Common mistakes

  • Pitching instead of listening. You should talk less than the customer.
  • Counting compliments. "Cool idea" is not demand. A prepayment, a deposit or a booked follow-up is.
  • Asking friends. They are not your ideal customer.
  • Stopping after launch. Talk to customers all the way to product-market fit and beyond.

For small teams

Conversations cost nothing but time, which makes this the best deal a bootstrapped founder has. Do 20 before you write code, then keep doing five a week.

Related terms

Sources

Back to the SaaS glossary