DEV Community

Alan Descure
Alan Descure

Posted on Originally published at smtm-store.pages.dev

Opportunity Discovery: A Practical Guide for When You're Stuck Looking for Problems Worth Solving

Opportunity Discovery: A Practical Guide for When You're Stuck Looking for Problems Worth Solving

The Problem in Concrete Terms

You're sitting at your desk with a notebook open, or staring at a blank Notion page, trying to find "a good opportunity." Maybe you've been laid off and want to build something. Maybe you're a PM tired of executing someone else's roadmap. Maybe you've got savings and six months of runway. The specific trigger doesn't matter—what matters is the loop you're stuck in:

You consume content about "finding opportunities" but produce zero validated leads.

Concretely, this looks like:

  • Reading 20 "problems worth solving" Twitter threads and saving none
  • Joining three Discord communities, lurking for weeks, posting nothing
  • Building a "pain point tracker" spreadsheet with 47 rows you never follow up on
  • Having five coffee chats where you ask "what's a problem you have?" and get vague answers you don't know how to evaluate
  • Cycling between "this idea is too small" and "this idea is too competitive" until you paralyze yourself

The output you actually need: a short list (3-5) of specific, validated problems where you can articulate the user, the pain, the current workaround, and why you're positioned to solve it. Everything else is theater.


Why the Usual Fixes Fail

1. "Scratch Your Own Itch" Is Incomplete Advice

It works if your itch is shared by thousands of people with budget. It fails when your itch is niche, unpaid, or already solved. Most people don't know how to test whether their personal pain generalizes.

2. "Talk to Customers" Without a Protocol

Cold outreach gets ignored. Warm intros run dry. Generic "can I pick your brain?" requests waste everyone's time. Without a repeatable interview structure, you collect anecdotes, not evidence.

3. Market Research Reports and Trend Lists

Gartner quadrants, "top 10 SaaS ideas 2024" newsletters, and keyword research tools show you where the crowd already is. By the time a trend is legible in data, the easy entry points are gone.

4. Idea Generation Frameworks (SCAMPER, Jobs-to-be-Done, etc.)

These are analysis tools, not discovery tools. They help you expand a seed, not find one. Applied to a blank page, they produce clever-sounding concepts with no tether to reality.

5. Waiting for "Founder-Market Fit" to Strike

Treating opportunity discovery as a lightning strike rather than a research process means you optimize for inspiration over evidence. You'll wait a long time.


Step-by-Step Solution

Phase 1: Define Your Constraints (Day 1)

Before hunting problems, define the box you're searching in. Write answers to these—don't just think them:

Capital & Runway

  • How many months can you work without revenue?
  • What's your monthly burn if you go full-time?
  • Do you need to pay yourself, or can you reinvest everything?

Time & Capacity

  • How many hours/week can you commit for the next 90 days?
  • Are you building solo, with a cofounder, or hiring contractors?
  • What technical depth can you actually ship? (Be honest: "I can build a React frontend and Supabase backend" ≠ "I can build anything")

Distribution Leverage

  • What audiences do you already have access to? (Newsletter, Twitter followers, former coworkers, industry Slack groups, conference circuits)
  • What channels have you actually used to get signups/sales before?
  • Which communities do you genuinely participate in, not just lurk?

Dealbreakers

  • Industries you won't touch (regulation, ethics, boredom)
  • Business models you won't run (marketplaces, hardware, ad-supported, enterprise sales >6 months)
  • Team structures you won't accept (solo, cofounder required, VC-backed only)

Output: A one-page "Founder Spec" document. Reference it every time you evaluate a lead. If it violates a constraint, discard it immediately—no "but maybe I could..." negotiations.


Phase 2: Mine Your Proximity (Days 2-7)

The highest-signal opportunities are adjacent to what you already know. Do these three exercises in order:

Exercise A: The Work History Audit (2 hours)

List every role you've held for 6+ months. For each, write:

  • Workflow: What process did you execute repeatedly?
  • Tools: What software did you use daily? What did you hate about it?
  • Workarounds: What did you build in Excel/Notion/Sheets/scripts because the tools failed?
  • Complaints: What did coworkers vent about at lunch or in Slack?
  • Budget: Who paid for tools? How much? Who approved purchases?

Don't filter. A junior PM's spreadsheet hack for tracking Jira dependencies is a better lead than a CTO's "AI strategy" vision.

Exercise B: The Vendor & Buyer Map (1 hour)

For each role, list:

  • Vendors you paid: Name, price, contract length, switching pain
  • Vendors you evaluated but didn't buy: Why not?
  • Internal tools your team built: What did they replace? What broke?
  • Procurement process: Who signed? Legal review? Security review? Pilot phase?

This maps the buying dynamics you'll eventually navigate.

Exercise C: The "Recent Friction" Log (Ongoing, 1 week)

Carry a note-taking tool. Every time you feel friction in any professional context—yours or observed—log:

  • Trigger: What were you trying to do?
  • Blocker: What stopped or slowed you?
  • Workaround: What did you do instead?
  • Frequency: How often does this happen?
  • Cost: Time lost, money spent, risk introduced, frustration level (1-5)
  • Actor: Who felt it? (Role, seniority, team size)

Target: 30+ entries. Don't judge. A designer complaining about Figma component library drift is as valid as a CFO complaining about SaaS sprawl.


Phase 3: Cluster & Filter (Days 8-10)

Take your 30+ friction entries. Group them by problem theme, not by tool or role. Example clusters:

  • "Engineering onboarding takes 3+ weeks because docs are stale"
  • "Marketing can't attribute leads to campaigns without manual UTM cleanup"
  • "Customer success lacks visibility into product usage pre-churn"
  • "Finance reconciles Stripe payouts manually every month"

For each cluster, create a Problem Card with:

  1. Problem statement (one sentence, no solution language)
  2. Who feels it (role, company stage, team size)
  3. Current workaround (specific tools + manual steps)
  4. Frequency & intensity (from your log)
  5. Budget owner (who can say yes to a purchase)
  6. Your proximity (how many degrees of separation from the buyer)
  7. Constraint check (pass/fail against your Founder Spec)

Filter ruthlessly. Keep only clusters that:

  • Score 4+ on frequency/intensity
  • Have a clear budget owner you can reach within 2 hops
  • Pass your Founder Spec constraints
  • You understand well enough to describe the workflow in detail

Target: 5-8 Problem Cards.


Phase 4: Validate with Structured Conversations (Days 11-30)

This is where most people stall. They either don't reach out, or they have bad conversations. Use this protocol:

The Outreach Template

Subject: Quick question about [specific workflow from their role]

Hi [Name],

I'm researching how [role] at [company stage/size] handle [specific problem]. You came up because [specific reason: former coworker, posted about X, spoke at Y].

Not selling anything. I have 3 specific questions that take ~10 minutes. Happy to share what I'm learning across all conversations.

Worth a quick call?

— [Your name], [one-line credibility: "ex-Stripe PM," "built internal tools at 200-person SaaS," etc.]

Send 20-30 of these. Expect 5-8 replies. Book 30-minute calls.

The Interview Script (Follow This Order)

  1. Context (2 min): "Walk me through the last time you did [workflow]."
  2. Pain (5 min): "What's the worst part? Show me the workaround if you can."
  3. Frequency (2 min): "How often does this come up? Last time it happened?"

Start here

Top comments (0)