Live Dec 7 & 14 · Free founding cohort Give your idea two Mondays ↗
The Builder’s Journal
GETTING STARTED · 3 MIN READ

How to start building a product when you have never built one before

A practical first-week plan: choose one person, find one recurring problem, and make the smallest useful thing you can test.

Start with access to a problem

Your first product does not need to be an app. It might be a template, a paid research brief, a repeatable service, a calculator, or a small software tool. Start with a problem you can observe and people you can reach. That gives you a way to learn before you spend heavily.

Write three moments from work or daily life when someone repeats a task, loses information, waits for an answer, or makes an avoidable mistake. Pick the one you understand best. A familiar, unglamorous problem is often a better learning project than a huge idea for an audience you cannot contact.

Day one: write a one-sentence promise

Use this format: For [person] who needs to [job], this product helps them [observable result] without [current friction]. For example: for freelance designers collecting feedback, this tool turns scattered comments into an editable revision list.

Underline everything you have not verified. Do designers actually lose feedback? Is a revision list the result they want? Do they already solve it well? Those uncertainties become your research questions. Download the problem brief to capture them.

Days two and three: learn how people do it now

Invite a small first batch of relevant people to describe the last time the problem happened. Ask what triggered it, what they tried, how long it took, and what happened when it went wrong. Ask to see the existing process when appropriate.

Record actual behavior separately from opinions. A person giving you an example to work on is a stronger signal than a person saying your idea sounds nice. A few conversations can uncover patterns, but they do not establish the size of a market.

Day four: choose the cheapest useful format

  • Knowledge problem: try a worked example, worksheet, or guide.
  • Execution problem: offer a small manual service or concierge pilot.
  • Calculation problem: start with a spreadsheet or simple calculator.
  • Repeated workflow problem: test a clickable prototype or one software flow.

For the feedback example, manually creating three revision lists may teach you more than building accounts and subscriptions. Use a document to prove the format is useful before automating it.

Day five: make the whole small experience

Include the instructions, the input, the result, and the next step. A beautiful screen with no usable outcome is an incomplete test. Define finished in plain language: a designer can paste sample comments, correct the proposed list, and copy the result.

If you use AI, give it that acceptance condition and one clear task. Ask it to identify assumptions before adding features. Keep an explicit later list so good ideas do not expand your first version.

Days six and seven: observe, then decide

Watch someone try the product without talking them through it. Ask what they expected at the moment they hesitate. Track whether they finish the task, whether the result is useful, and whether they would take a concrete next step.

Choose one of three moves: continue because behavior supports the idea; change because a specific obstacle is visible; or pause because the problem is not important enough. Write the reason down. Learning that a feature is unnecessary is progress too.

Your next thirty minutes

List one audience, one problem, and one person you can ask about it. Then write a short invitation to learn about their current process. You can draft with AI, but use your own judgment before sending it.

Need structure? Follow the free twelve-stage learning path, work through Waymaker’s approach, or bring your idea to the Two Mondays workshop.

Make your next step a real one.

Use a free worksheet to define your product, or explore the Two Mondays workshop for time to build with Ashley.