Pre-PMF User Discovery: Find Users Before You Build

Most founders waste months trying to find users after building. Learn a proven framework for discovering users before you know what you're building - works for pre-PMF startups.

I've watched hundreds of founders make the same mistake: they build a product, then spend months trying to find people who want it. They post on Product Hunt, run ads, cold email, and hope something sticks.

But here's what I've learned after building multiple products and talking to successful founders: early user discovery shouldn't happen after you build. It should happen before you know what you're building.

The Problem with Traditional User Discovery

Most founders think user discovery means:

1. Build a product
2. Create a landing page
3. Post on social media
4. Run ads
5. Hope people sign up

This approach has three fatal flaws:

Flaw 1: You're guessing what people want You built based on assumptions, not evidence. You don't know if anyone actually needs this.

Flaw 2: You're competing for attention By the time you launch, you're competing with established products, ads, and noise. Getting noticed is expensive and hard.

Flaw 3: You're too late If you're trying to find users after building, you've already invested time and money in something that might not work.

The Better Way: Problem-First Discovery

Instead of building first and finding users later, flip the process:

1. Find people with problems (not people who might want your solution)
2. Understand their pain deeply (what they're using now, why it sucks, what they wish existed)
3. Engage before building (get commitments, validate willingness to pay)
4. Build only what they ask for (not what you think they need)

This isn't theoretical. This is how successful founders actually do it.

Where Your First Users Actually Are

Your first users aren't on Product Hunt or Twitter waiting to discover your product. They're on Reddit, Hacker News, Stack Overflow, GitHub, and niche forums, actively discussing problems and looking for solutions.

Here's where to look:

Reddit: The Honest Problem Forum

Reddit is where people go to complain, ask questions, and look for help. It's unfiltered and honest.

What to search for:

• "Is there a tool that..."
• "I'm tired of using..."
• "Does anyone know a better way to..."
• "[Competitor] is too expensive, alternatives?"

Example: A founder building a simple CRM found 30+ Reddit threads where people said:

"I just need something simple. Salesforce is overkill, HubSpot is too complex, and spreadsheets are a nightmare. Why isn't there a simple CRM for small teams?"

That's not just feedback - that's a validated market with clear requirements.

Hacker News: Where Early Adopters Live

Hacker News is full of technical founders and early adopters who love trying new tools.

What to look for:

• "Show HN" posts in your space (see what people ask for)
• Comments on competitor launches (what's missing?)
• "Ask HN" threads about problems you solve
• Discussions about tool alternatives

Stack Overflow & GitHub: Developer Pain Points

If you're building developer tools, these platforms are goldmines.

Stack Overflow: Questions reveal what developers struggle with GitHub Issues: Feature requests show what's missing from existing tools

Long-form Q&A (manual)

Sites like Quora can still host explicit questions. Needle Search does not index Quora - use Needle for the supported communities, then spot-check Q&A sites by hand when it helps.

The Pre-PMF Discovery Framework

Here's the exact framework I use and recommend to founders:

Phase 1: Problem Discovery (Week 1-2)

Goal: Find 50+ people actively discussing a problem you might solve.

Steps:

1. List 5 problems you're considering solving

2. Search each problem across Reddit, HN, Stack Overflow, and the other communities your Needle plan includes

3. Count the signal:

   • How many threads discuss this?
   • How recent are they?
   • How much engagement (upvotes, comments)?
   • What's the emotional tone (frustrated vs. curious)?

4. Pick the problem with the strongest signal

Example: I was considering three ideas:

• Idea A: 12 threads, low engagement → Skip
• Idea B: 47 threads, high engagement, recent → Explore
• Idea C: 8 threads, mostly old → Skip

Idea B became Needle.

Phase 2: Deep Understanding (Week 2-3)

Goal: Understand the problem deeply before building anything.

Steps:

1. Read 20+ threads about the problem

2. Identify patterns:

   • What are people currently using?
   • Why does it suck?
   • What would an ideal solution look like?
   • What are they willing to pay?

3. Create a problem statement:

   "[Target user] is frustrated with [current solution] because [specific pain]. They need [what they're asking for] but current tools are [why they don't work]."

Example problem statement:

"Solo founders are frustrated with manually searching Reddit, Hacker News, and Stack Overflow to find customers because it takes 10+ hours per week. They need a tool that searches the communities they care about in one pass, but current solutions are either too expensive ($100+/month) or too complex (require API setup)."

Phase 3: Engagement & Validation (Week 3-4)

Goal: Get 10+ people to commit to trying your solution.

Steps:

1. Engage in 10-15 relevant discussions
2. Provide value first (answer questions, share insights)
3. Mention you're building something (casually, not salesy)
4. Ask: "Would you try it if I built it?"
5. Get email addresses or commitments

Engagement template:

"I've been thinking about this problem too. I'm building a simple tool that does [specific thing]. Would you be interested in trying it when it's ready? What features would be must-haves for you?"

Success criteria:

• 10+ people say "yes, I'd try it"
• 5+ people give specific feature requests
• 3+ people ask "when will it be ready?"

If you can't hit these numbers, the problem might not be urgent enough.

Phase 4: Build & Iterate (Week 4+)

Goal: Build only what validated users asked for.

Steps:

1. Build the minimum version (only must-have features)
2. Share with your validated users first
3. Get feedback and iterate
4. Only then, expand to broader marketing

Real Example: How a Founder Got 73 Beta Users in 10 Days

A founder I know was building an AI meeting summarizer. Instead of building first, he:

1. Searched Reddit and Stack Overflow for "meeting notes problem"
2. Found 40+ threads where people complained about taking notes
3. Engaged in 15 discussions, asking what they'd want in a solution
4. Got 23 people to commit to trying it
5. Built a simple MVP based on their specific requests
6. Launched to those 23 people first
7. Got 73 signups in 10 days (they shared it with others)

Total ad spend: $0. Total time to first users: 2 weeks.

Compare that to the traditional approach: build for 3 months, launch on Product Hunt, get 50 signups, 3 convert, realize you built the wrong thing.

Tools That Make This Faster

Manually searching platforms works, but it's slow. Here are tools that help:

Social Listening Tools:

• Search multiple supported communities in one workflow
• Track conversations over time
• Filter by sentiment and engagement
• Highlight competitor names in posts when you save competitors (on eligible plans)

My recommendation: Use a tool like Needle so Search covers 10+ communities on your plan (see Pricing), with Auto Search for your brand and optional digests. For a complete guide on how to find your first 100 customers, check out our comprehensive strategy guide.

Why it matters: Instead of spending 10 hours per week manually searching, you can find the same opportunities in 30 minutes.

Common Mistakes Pre-PMF Founders Make

Mistake 1: Building before validating You're excited about your idea, so you build it. But excitement ≠ market demand.

Mistake 2: Asking "would you use this?" People lie. Instead, ask "what are you currently using and why does it suck?"

Mistake 3: Only looking on one platform Your users might be on Reddit, not Twitter. Search everywhere and use Auto Search or digests to stay on top. Learn more about finding your first 100 customers across multiple platforms.

Mistake 4: Giving up too early You search for 2 hours, find nothing, assume there's no market. But maybe you're searching wrong, or the problem is discussed differently than you think.

Mistake 5: Not engaging Reading threads isn't enough. You need to talk to people, ask questions, get commitments.

The Pre-PMF Discovery Checklist

Before building, make sure you can answer:

• Can I find 30+ people actively discussing this problem?
• Do they mention what's missing from current solutions?
• Are they frustrated enough to try something new?
• Can I get 10+ people to commit to trying my solution?
• Do they mention what they'd pay (or are they only looking for free)?
• Is there a clear gap I can fill (not just "I can do it better")?

If you can't check all boxes, keep discovering. It's cheaper to spend a month finding the right problem than to spend 6 months building the wrong solution.

Final Thoughts

Early user discovery isn't about marketing your product. It's about finding problems worth solving.

The founders who succeed aren't the ones who build the best products. They're the ones who find the right problems to solve.

Your next steps:

1. Stop building (if you haven't started) or pause (if you have)
2. List 5 problems you're considering solving
3. Search each one across Reddit, HN, Stack Overflow, and the other communities your Needle plan includes
4. Pick the one with the strongest signal
5. Engage with 10-15 people discussing it (learn how to validate your startup idea with real conversations)
6. Get commitments before building
7. Build only what they ask for

Remember: It's better to discover for a month than to build for 6 months and realize nobody wants it.

The best time to find your first users is before you know what you're building. The second best time is now.