Most side projects don’t fail because the code is bad. They fail because nobody needed them, or because the founder never asked anyone to pay. For engineers selling into English-speaking markets, there’s an extra barrier: the conversations feel harder than the code, so they get postponed. This guide is about having those conversations early, before you build more.
01Start with the customer, not the product
A side project usually starts with “I could build this”. A business starts with “these people have this problem and already spend time or money on it”. Before writing more code, write one sentence:
“[Who] struggles with [problem], and today they [current workaround].”
If you can’t fill in the last part, you don’t know the market yet. That’s fine; it’s what the next step is for.
02Customer conversations that tell you the truth
People are polite. Ask them about your idea and they’ll say it’s great. Ask about their past behaviour and you learn what they actually do. This is the core idea of The Mom Test by Rob Fitzpatrick, a short book many founders use.
- Talk about their life, not your idea. Save the product for the end, if at all.
- Ask about specific times: “the last time”, not “usually”.
- Listen for money and time already spent. That’s the strongest sign of a real problem.
- Write down exact words people use. They become your website copy later.
03Find your first ten users by hand
In the beginning, growth comes from doing things that don’t scale, as Paul Graham argues in his well-known essay: writing to people one by one, helping them set up, calling them afterwards.
Look where your users already talk about the problem: specific subreddits, Slack and Discord communities, LinkedIn posts, GitHub issues of tools they use. Twenty personal messages will teach you more than a launch post.
04Ask for money early
Compliments and sign-ups don’t prove much. Payment does. Ask for it sooner than feels comfortable:
- Offer a paid pilot to a small group: a fixed price, a fixed period, your personal help.
- Say the price clearly and stop talking. “Is that okay for you?” invites a discount.
- Treat “no” as information: ask what would make it worth paying for.
- Don’t start free “for now”. It’s much harder to start charging later than to give a discount.
05Write a page that says what you sell
Engineers often describe how the product works. Customers want to know what it does for them. Lead with the outcome and the audience:
One page, one clear headline, one button. Use your customers’ own words from the interviews; they convert better than your words.
06A weekly rhythm alongside your job
- Protect a few fixed hours each week, and spend at least half of them talking to users, not coding.
- Track one number that matters: conversations held, paying customers, or weekly active users.
- Check your employment contract for clauses about side work and intellectual property before you charge anyone. If you’re unsure, get advice.
- Decide in advance what result after three months means “continue”, “change” or “stop”.
07Phrases for customer conversations
| Situation | Phrases |
|---|---|
| Learning about the problem | “Tell me about the last time…” · “How do you deal with that today?” · “What have you already tried?” · “What did that cost you?” |
| Digging deeper | “Interesting. Why was that a problem?” · “Can you show me how you do it now?” · “Who else is involved when that happens?” |
| Keeping it a conversation, not a pitch | “I’m not selling anything yet. I’m trying to understand the problem.” · “Is there anyone else I should talk to about this?” |
| Asking for money | “We’re starting a paid pilot with five teams at 99 dollars a month. Would you like to be one of them?” · “What would make this worth paying for?” |
| Learning from new users | “What almost stopped you from signing up?” · “What would you miss if this disappeared tomorrow?” |
| Reacting to criticism | “Thanks for the honest feedback.” · “That’s really useful. Can I come back to you when I’ve changed it?” |
| Instead of | Try |
|---|---|
| “Would you use this?” | “How do you handle this today?” |
| “It’s an AI-powered platform.” | “It tells you which service is driving your bill.” |
| “Do you like my idea?” | “What would stop you from using it?” |
| “It’s free for now.” | “The pilot is 99 dollars a month, and you can cancel anytime.” |
| “Sorry to bother you…” | “I saw your post about X and had one question.” |
08Example: from first message to first call
An invented example. The email is short, personal and honest about not selling yet; the first question on the call is about the past.
Outreach email and call opening
Subject: Your post about AWS costs
Hi Laura,
I saw your post about your AWS bill jumping after the migration. I’m an engineer researching how small teams keep cloud costs under control, and your experience sounds very relevant.
Would you be open to a 20-minute call next week? I’m not selling anything yet. I’d just like to hear how you deal with it today. Happy to share what I learn from other teams.
Thanks,
Tomás
Laura: [on the call] So, what would you like to know?
Tomás: Thanks for making time. Could you tell me about the last time the bill surprised you? What happened, step by step?
09Before you build the next feature
- I can say who has the problem and what they do about it today.
- I’ve had at least ten conversations about their past behaviour.
- I’ve written down the exact words customers use.
- I’ve asked at least three people to pay.
- My page leads with the outcome and the audience.
- I’ve checked my employment contract.
- I know my one number and my three-month decision point.