User Feedback for AI Coders
The cheapest, fastest way to make your app better, and you're probably ignoring it.
In this chapter
- Why user feedback is the single most undervalued thing in AI coding
- Five simple ways to collect feedback (no code needed)
- How to ask the right questions so the answers are actually useful
- What to do with feedback once you have it
Why You Need Feedback
When you build something on your own, you fall in love with it. Every button is in the "obvious" place. Every word makes perfect sense. Every feature is exactly what people want.
Then you show it to a real person and they immediately get stuck on something you never noticed. That moment is worth more than a week of solo work. It tells you what to fix.
5 Simple Ways to Get Feedback
You don't need fancy tools. Pick one of these and start today.
Asking the Right Questions
The biggest mistake is asking questions that have a built-in answer. "Do you like my app?" is useless: they'll say yes to be polite. The trick is asking open questions that get real stories.
❌ Useless questions
- "Do you like it?"
- "Is this useful?"
- "Would you use it again?"
- "Any feedback?"
- "Was it easy?"
✅ Useful questions
- "What were you trying to do when you opened it?"
- "What was the most confusing part?"
- "What did you expect to happen that didn't?"
- "What would make this a no-brainer for you?"
- "Walk me through how you'd use this for your work."
The Mom Test: 3 Rules
There's a famous book called The Mom Test by Rob Fitzpatrick. The whole idea is that if you ask the wrong way, even your mom will lie to you to be nice. Three rules from the book:
Your First Feedback Loop
You don't need a system or a tool. You need 5 conversations. Here's how to set them up this week:
-
1
Pick 5 people who match your target user: friends, coworkers, online communities, your email list. They don't have to be strangers.
-
2
Send them a short message: "I'm building a tool for [their problem]. Could I show it to you for 15 minutes and get your reaction? I'd love your honest take, even if it's 'this is bad.'"
-
3
Schedule a 15-minute call. Show them your app. Don't explain anything. Watch them try to use it.
-
4
Ask: "What were you trying to do? What was confusing? What would make this useful for you?"
-
5
Take notes. Look for patterns across all 5 people. The same complaint from 3 different people = your top priority fix.
What to Do With the Feedback
Collecting feedback is the easy part. Acting on it is where most people fail. Don't fall into these traps:
Building everything anyone asks for
Wait until 3+ people ask for the same thing. One-off requests usually come from edge cases. Patterns matter, single voices don't.
Defending your design when they don't like it
Don't argue. Don't explain why it's "actually" right. Just say "Tell me more." You're trying to learn, not win.
Ignoring negative feedback because it stings
The hardest feedback is the most valuable. The people who care enough to complain are the ones who care enough to come back if you fix it.
Letting feedback pile up unread
Every Friday afternoon, spend 30 minutes reading the week's feedback. Sort into bugs, ideas, and vibes. Pick the top 3 things and add them to next week's work.
Real World Example: The 5-call rule before launch
Before you spend another month polishing your app, do this: find 5 real people in your target audience and watch each of them try to use your app for 15 minutes. You'll find more problems in those 75 minutes than in three months of solo work, and you'll fix the right ones, not the ones you imagined existed.
“I want to add a small feedback widget to my app: a floating button bottom-right that opens a simple form. The form should ask one question: "What were you trying to do?" with a textarea, plus an optional email field if they want a reply. Save submissions somewhere I can read later (use the simplest option you can, like a database table or even an email to me). Don't over-engineer it.”
Prompt This
“Add a small feedback button to my app, in the bottom-right corner. When clicked, it opens a tiny form asking "What's on your mind?" with a text area and a Send button. After they hit Send, save the message somewhere I can read it later and show a friendly "Thanks!" message. Don't make it complicated: keep it as simple as possible.”
“Whenever someone sends feedback through my app, automatically email me the message at [your email]. Include the time they sent it and (if they're logged in) their name and email. That way I see new feedback the moment it comes in instead of having to check a dashboard.”
“Right after a new user signs up to my app, show them a one-question survey: "What are you hoping to do with [my app]?" with a text area. Save their answer with their account so I can read it later. Make it skippable: don't force anyone to answer.”
Frequently Asked Questions
How early should I start asking for feedback?
From day one. Even with one or two friends trying your app, ask them what felt confusing. The earlier you find problems, the cheaper they are to fix. Waiting until you have hundreds of users means you've already built the wrong thing for hundreds of people.
What if people say my idea is bad?
Lucky you: you found out before spending a year on it. Don't take it personally. Negative feedback isn't an attack, it's a gift. Ask follow-up questions: 'What part felt wrong?' and 'What would you want instead?' Those answers are gold.
How do I get strangers to give me feedback?
Make it embarrassingly easy. A 1-question survey. A 'Reply to this email and tell me what you think' line. A button on your site that says 'Send feedback' and opens a tiny form. The lower the friction, the more people respond.
Should I give people something in exchange for feedback?
For longer feedback (15-30 minute calls), yes: a $10 gift card or free month of your product is fair. For quick replies and short surveys, a heartfelt thank-you is enough. Don't make it transactional or you'll attract people who only want the reward.