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

Opportunity Discovery: A Practical Guide for When You're Stuck

The Problem in Concrete Terms

You're sitting at your desk with a notebook full of half-baked ideas, or worse—a blank page. You want to build something people will pay for, but every direction feels either obvious (already saturated) or obscure (no market). You've read about "finding your niche" and "scratching your own itch," but your itches are either too small to sustain a business or too weird to explain to anyone else.

The stuckness shows up in specific ways:

  • You've interviewed five potential customers and heard five completely different problems. No pattern emerges.
  • You've built three MVPs that nobody used. Each time you were convinced this was the one.
  • You spend weeks researching a market, find a genuine gap, then discover three well-funded competitors you missed.
  • You have a full-time job and 10 hours a week. You can't afford to chase dead ends.
  • Every framework you try—Jobs to Be Done, Value Proposition Canvas, Blue Ocean—feels like corporate theater when you actually sit down to use it.

The core issue: opportunity discovery is treated as a creative exercise when it's actually a research discipline. You're waiting for inspiration to strike instead of running a systematic process that produces signal.


Why the Usual Fixes Fail

1. "Talk to customers" without a hypothesis.

You book calls, ask open-ended questions, and get polite, unfocused answers. People don't know what they want built; they know what frustrates them right now. Without a specific hypothesis to test, conversations produce noise, not signal.

2. "Scratch your own itch" when your itch isn't representative.

You're a developer who hates configuring CI/CD pipelines. You build a tool for that. Turns out most developers tolerate the pain, and the ones who don't already use GitHub Actions. Your problem is real but not commercially real.

3. Keyword research and SEO tools as opportunity radar.

Ahrefs tells you "best project management software" has 40K monthly searches. It doesn't tell you which project management workflows are underserved, or whether the searchers have budget authority. Search volume ≠ willingness to pay.

4. Trend-chasing (AI, Web3, no-code, whatever's next).

By the time a trend shows up in TechCrunch, the low-hanging fruit is gone. You're competing with funded teams who started six months ago. Trends are useful context, not starting points.

5. Frameworks as substitutes for judgment.

Filling out a Value Proposition Canvas doesn't validate anything. It organizes your assumptions. The hard work—falsifying those assumptions—happens outside the template.


Step-by-Step Solution

Phase 1: Define Your Constraints (Week 1)

Before looking outward, lock down your inward reality. Write answers to these questions in a single document:

Constraint Your Answer
Hours/week available
Runway (months without revenue)
Skills you can apply without learning new ones
Distribution channels you already have access to (audience, network, newsletter, SEO assets)
Price floor: minimum monthly revenue to make this worth continuing
Deal-breakers (industries you won't touch, business models you hate, etc.)

Why this matters: A "great opportunity" requiring 40 hours/week, $50K upfront, and enterprise sales experience is a bad opportunity for you. Constraints filter 90% of noise before you start.

Action: Complete the table. Share it with one person who knows you well. Ask: "Does this feel honest?"


Phase 2: Map Adjacent Problems (Weeks 1–2)

Don't hunt for "ideas." Hunt for expensive, recurring problems in domains you already understand.

Step 2.1: List your "earned secrets."

What do you know deeply because you've lived it? Not "I read about X" but "I spent 2,000 hours doing X." Examples:

  • Managing vendor onboarding at a mid-size hospital
  • Scaling a Discord community from 0 to 5,000 active members
  • Debugging memory leaks in React Native at 3 AM
  • Negotiating SaaS renewals for a 200-person company

Step 2.2: For each earned secret, extract 5–10 workflows.

Break the domain into discrete, repeatable workflows. Not "hospital operations" but:

  • Credential verification for traveling nurses
  • Shift-swap approval chains
  • Equipment maintenance request routing
  • Compliance audit evidence collection

Step 2.3: Score each workflow on three dimensions (1–5 scale).

Workflow Frequency (how often) Pain Intensity (how much it sucks) Budget Visibility (who pays, how easy to reach)
Credential verification 5 4 3
Shift-swap approval 4 3 2
Equipment maintenance 3 5 4
Compliance audit prep 2 5 5

Step 2.4: Keep only workflows scoring ≥11 total.

Discard the rest. You now have a shortlist of specific, high-pain, budget-visible problems in domains you actually understand.

Action: Produce a ranked list of 3–5 workflows. Time box: 10 hours max.


Phase 3: Validate the Problem Exists (Weeks 2–4)

For each shortlisted workflow, run a Problem Validation Sprint. Goal: confirm the problem is real, urgent, and currently unsolved well enough.

Step 3.1: Write a Problem Hypothesis (one sentence).

Template: "[Specific role] spends [time/money] on [workflow] because [root cause], and current solutions fail to [specific gap]."

Example: "Travel nurse coordinators spend 15+ hours/week manually verifying licenses across 12+ state portals because each state has a different API/format, and current VMS tools only cover 60% of states."

Step 3.2: Identify 20 people who currently do this workflow.

Not "potential users." People doing it today. Sources:

  • LinkedIn: search by title + company size + keywords from your workflow
  • Industry Slack/Discord communities (lurk first, then ask)
  • Conference attendee lists (often public)
  • Your own network—ask "Who handles X at your company?"

Step 3.3: Run 10-minute discovery calls.

Script:

  1. "I'm researching [workflow]. Can you walk me through the last time you did it?" (Listen. Take notes. Don't pitch.)
  2. "What tools or workarounds are you using?"
  3. "What's the worst part?"
  4. "If you could wave a wand and fix one thing, what would it be?"
  5. "Who else should I talk to?"

Step 3.4: Score each call.

After each call, rate:

  • Problem confirmed? (Y/N) — Did they describe the exact pain without prompting?
  • Urgency? (1–5) — Are they actively looking for a fix now?
  • Budget signal? (Y/N) — Did they mention paying for tools, contractors, or overtime to handle this?
  • Workaround complexity? (1–5) — How messy is their current solution? (Spreadsheets + Zapier + manual steps = high score)

Step 3.5: Decision rule.

After 5 calls per workflow:

  • ≥3 "Problem confirmed = Y" + ≥3 "Urgency ≥4" + ≥2 "Budget signal = Y" → Proceed to solution validation
  • Otherwise → Kill the workflow. Move to next.

Action: Complete validation for your top 3 workflows. Expect 15–20 calls total. Time box: 2 weeks.


Phase 4: Validate Willingness to Pay (Weeks 4–6)

Problem existence ≠ commercial viability. Now test if they'll pay you to solve it.

Step 4.1: Create a "concierge MVP" offer.

Not software. A manual service that delivers the outcome. Examples:

  • "I'll verify your next 50 nurse licenses in 48 hours for $500."
  • "I'll run your compliance audit prep and deliver the evidence package for $2,000."
  • "I'll build and maintain your shift-swap automation in Airtable for $300/mo."

Price at ~1/10th the cost of their current

Start here

Top comments (0)