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 sit down to "find a business idea" or "discover an opportunity" and end up doing one of three things: scrolling Twitter for inspiration, browsing Product Hunt hoping something sparks, or staring at a blank notebook waiting for lightning to strike. None of it works. The ideas you generate feel either too small (a slightly better to-do app), too vague ("AI for healthcare"), or already saturated (the tenth newsletter curation tool).

The concrete reality: you lack a repeatable process for finding problems that real people urgently need solved and will pay for. You're confusing idea generation with opportunity discovery. Ideas are cheap. Opportunities are specific, validated intersections of painful problems, reachable audiences, and feasible solutions.

You know this intellectually. But when you try to act on it, you hit walls:

  • You talk to five people, get polite nods, and learn nothing actionable
  • You research a market for weeks, compile a spreadsheet, and still can't decide if it's worth pursuing
  • You build an MVP for a problem you think exists, launch to crickets, and realize you solved a "nice to have" not a "must have"
  • You keep abandoning directions because you can't tell signal from noise

The stuckness isn't creative. It's methodological.


Why the Usual Fixes Fail

1. "Scratch your own itch"

Works if your itch is shared by thousands of people with budget authority. Fails when your problem is idiosyncratic, too niche, or already solved well enough that switching costs block adoption. Most developers overestimate how common their workflow frustrations are.

2. "Talk to customers"

Generic advice. Without a framework for who to talk to, what to ask, and how to interpret answers, you get surface-level feature requests ("I wish this had dark mode") instead of discovered problems ("I lose three hours a week reconciling these two systems manually").

3. "Look for complaints on Reddit/Twitter/forums"

Complaints are noise until filtered. People complain about everything. The signal—recurring, expensive, urgent problems tied to workflows where money already changes hands—gets buried in venting, feature wishes, and jokes.

4. "Analyze market trends and TAM"

Top-down analysis creates false confidence. A $50B market with 500 competitors and high switching costs is worse than a $50M market with 3 competitors and desperate buyers. Spreadsheets don't reveal purchasing urgency.

5. "Build in public and see what resonates"

This tests marketing, not opportunity. You can build an audience for the wrong problem. Validation requires evidence of willingness to pay, not willingness to follow.


Step-by-Step Solution: The Opportunity Discovery Loop

This process takes 2–4 weeks part-time. It produces a shortlist of 1–3 opportunities with evidence, not hunches.

Phase 1: Define Your Search Boundaries (Days 1–2)

Don't search "everywhere." Constrain by what you can actually execute.

Write down your constraints explicitly:

  • Technical: What can you build or credibly partner to build? (e.g., "I can build B2B SaaS, not hardware or regulated fintech")
  • Distribution: What channels do you have or can access? (e.g., "I know 20 VP Engineering types, zero dental practice owners")
  • Capital: How long can you go without revenue? (e.g., "6 months runway, then need $5k MRR")
  • Interest: What domains will you still care about in 18 months? (List 3–5, exclude the rest)

Output: A one-page "Opportunity Filter" document. Every candidate idea gets scored against this. If it fails a hard constraint, discard it immediately.


Phase 2: Map Money Flows, Not Problems (Days 3–7)

Problems are everywhere. Paid problems cluster where money already moves.

Pick 3–5 verticals that pass your filter. For each, answer:

  1. Who pays whom, for what, how often, and why?

    Trace 5–10 real transactions. Not "companies buy software." Specifics: "Series A B2B startups pay $2–5k/mo for sales engagement tools because they need predictable pipeline and their CRM doesn't automate follow-up."

  2. Where is the friction in that transaction?

    Look for: manual workarounds, spreadsheets bridging systems, hired contractors doing repeatable work, complaints about speed/accuracy/compliance, homegrown internal tools.

  3. What triggers a purchase?

    New hire? Audit? Funding round? Regulation change? Seasonal spike? Trigger = urgency.

Method:

  • Find 3 people per vertical who buy (not use) the relevant tools/services. Ask: "Walk me through the last time you bought [category]. What prompted it? What almost stopped you? What would make you switch?"
  • Read 20 job descriptions for the buyer role. Responsibilities = budget authority. Pain points hide in "must handle X" and "responsible for Y."
  • Scan 10 vendor pricing pages. Pricing tiers reveal what buyers value (seats? volume? features? compliance?).

Output: For each vertical, a one-pager: Buyer, Budget, Trigger, Current Solutions, Friction Points, Switching Barriers.


Phase 3: Run Problem Discovery Interviews (Days 8–18)

Not "customer discovery." Problem discovery. You're not validating a solution. You're hunting for problems that meet three criteria: Painful, Urgent, Paid.

Recruit 15–20 interviewees per promising vertical.

Sources (in priority order):

  1. Warm intros from your network to buyers (not users)
  2. Cold LinkedIn: "I'm researching [workflow], not selling. 15 min, I'll share findings."
  3. Industry Slacks/Discords: "Happy to buy coffee for 15 min on [topic]"
  4. Conference attendee lists (cold email, low conversion but high quality)

The Interview Script (15–20 min max):

  1. Context: "Walk me through your typical [workflow/process] from start to finish." (Listen for steps, tools, handoffs)
  2. Pain: "Where does it break, slow down, or require manual work?" (Get specifics: "Every Friday I spend 2 hours copying data from X to Y because they don't sync")
  3. Impact: "What happens if that doesn't get done? What's the cost?" (Quantify: lost revenue, compliance risk, missed SLA, overtime)
  4. Current Solution: "How do you handle it today? What have you tried?" (Reveals switching costs, workarounds, competitor landscape)
  5. Trigger: "When did you last look for a better way? What prompted it?" (Urgency signal)
  6. Budget: "If a tool solved this perfectly, who approves the spend? What's the process?" (Buying authority, procurement friction)
  7. Magic Wand: "If you could wave a wand and fix one thing about this workflow, what would it be?" (Unconstrained problem statement)

Critical rules:

  • Record (with permission). Take notes after, not during.
  • Zero pitching. Zero "would you use X?" questions.
  • If they mention a problem, ask "Tell me about the last time that happened." Specifics > generalizations.
  • Score each problem mentioned: Pain (1–5) × Urgency (1–5) × Budget Evidence (1–5). Keep only 15+ scores.

Output: A problem database (Notion/Airtable/Sheets) with: Verbatim quote, Pain/Urgency/Budget scores, Buyer role, Trigger, Current workaround, Switching barrier.


Phase 4: Stress-Test the Top 3 Problems (Days 19–25)

For each high-scoring problem, run three validation checks.

Check 1: The "Pay Now" Test

Create a one-pager: Problem, Proposed Solution (2 sentences), Price, Timeline. Send to 5–10 interviewees who mentioned this problem. Ask: "If this existed at this price on this timeline, what would stop you from buying?"

  • "Nothing" = false positive (politeness)
  • Specific objections (security, integration, procurement, timing) = real signal
  • "I'd need to see a demo" = request a paid pilot commitment ($500–2k for 3

Start here

Top comments (0)