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 have time, skills, and maybe even capital—but every idea you generate feels either obvious, saturated, or pointless. You've read the "scratch your own itch" advice, but your itches are either too niche (nobody else has them) or too broad (everyone's already building for them). You've tried browsing Product Hunt, Indie Hackers, and Reddit for "problems people complain about," but the complaints are either feature requests for existing tools or vague frustrations without clear solutions.

The reality: you're not lacking creativity. You're lacking a systematic way to observe, filter, and validate opportunities that doesn't rely on inspiration striking.


Why the Usual Fixes Fail

1. "Scratch your own itch" assumes your problems are representative. They rarely are. Your workflow, constraints, and tolerance for friction are shaped by your specific context. Building for yourself often produces tools only you understand.

2. Keyword research and trend reports show demand, not opportunity. High search volume means competition. Low search volume means no market. The sweet spot—real problems with insufficient solutions—doesn't announce itself in SEO tools.

3. Customer interviews without a hypothesis waste everyone's time. Asking "what's your biggest pain point?" gets you complaints about weather, traffic, and software that already exists. You need a specific lens before you talk to anyone.

4. Copying funded startups puts you 18 months behind. By the time a company raises Series A, the obvious version of their idea is taken. The opportunity is in the adjacent problems they're ignoring or the segments they're underserving.

5. Waiting for "founder-market fit" is passive. You don't discover fit by thinking. You discover it by testing specific hypotheses against reality.


Step-by-Step Solution

Phase 1: Build Your Observation Lens (Days 1–3)

1.1 List your unfair advantages
Write down every specific advantage you have: technical domains you know deeply, industries you've worked in, communities you're part of, regulatory environments you understand, distribution channels you access, hardware you own, data you can reach. Be specific. "I know Python" is useless. "I've built three internal tools for logistics teams at mid-size freight forwarders" is a lens.

1.2 Map your adjacency zones
For each advantage, list three adjacent domains. If you know freight forwarding, adjacent zones might be: last-mile delivery coordination, customs brokerage workflows, warehouse labor scheduling, cross-border returns processing. These are where you have enough context to evaluate problems but aren't already saturated.

1.3 Define your constraint envelope
Write down your hard constraints: maximum hours/week you'll invest, capital you'll risk, skills you won't learn (e.g., "no mobile development," "no sales-led enterprise"), timeline to first revenue. This prevents you from falling in love with opportunities you can't actually execute.

Phase 2: Systematic Problem Hunting (Days 4–14)

2.1 Run "workflow shadowing" sessions
Find 5–10 people in your adjacency zones. Ask to watch them work for 90 minutes via screen share or in person. Do not interview. Do not ask questions during. Take timestamped notes on:

  • Workarounds (spreadsheets, copy-paste, manual retyping)
  • Handoffs between tools or people
  • Moments of visible frustration or hesitation
  • Steps repeated more than 3x/day
  • Information hunted across multiple tabs/systems

2.2 Extract problem candidates
After each session, spend 15 minutes converting observations into problem statements using this template:

[Role] spends [time/frequency] doing [specific manual step] because [tool gap/process break], resulting in [quantifiable cost: errors, delays, rework, compliance risk].

Discard anything without a quantifiable cost. "It's annoying" isn't quantifiable. "3 hours/week reconciling shipping manifests" is.

2.3 Score each candidate
Rate each problem 1–5 on:

  • Frequency: How often does it occur? (Daily = 5, Monthly = 1)
  • Pain intensity: How costly is each occurrence? (Revenue loss/compliance risk = 5, Minor annoyance = 1)
  • Solution clarity: Can you envision a specific software/workflow fix? (Clear = 5, Vague = 1)
  • Access: Can you reach 10+ decision-makers in this role within 2 weeks? (Yes = 5, No = 1)
  • Constraint fit: Does it fit your envelope from 1.3? (Yes = 5, No = 1)

Keep only candidates scoring ≥18/25.

Phase 3: Rapid Validation (Days 15–30)

3.1 Build a "concierge MVP" for the top 2–3 candidates
Do not write code. Design a manual service that solves the problem:

  • Define the exact input the user provides
  • Define the exact output you deliver
  • Set a price (even $1) and a delivery timeline
  • Find 3–5 people to pay you for one cycle

Example: If the problem is "freight coordinators spend 3 hours/week manually matching invoices to bills of lading," your concierge MVP: they email you the PDFs Friday 5pm, you return a matched spreadsheet Monday 9am, charge $50/week.

3.2 Run the concierge cycle
Execute manually for 2–4 weeks. Track:

  • Time you spend per client
  • Error rate and rework requests
  • Whether they renew without prompting
  • What they ask you to add/change
  • Whether they refer colleagues

3.3 Decide: automate, pivot, or kill

  • Automate if: renewal rate >80%, you understand the pattern well enough to specify software, unit economics work at 10x volume.
  • Pivot if: they pay but use it differently than expected, or a adjacent problem emerges as more urgent.
  • Kill if: renewal <50%, economics don't scale, or you hate the work.

Phase 4: Transition to Build (Day 31+)

4.1 Write the spec from concierge learnings
Document every edge case you handled manually. Every template you created. Every decision rule you applied. This is your product spec—not a guess, but a record.

4.2 Build the thinnest slice that replaces you
Automate only the highest-frequency, highest-certainty step first. Keep human-in-the-loop for everything else. Ship to your concierge clients.

4.3 Expand outward only when pulled
Add features only when existing paying users ask for them specifically. Ignore "nice to have" feedback from non-users.


Checklist

  • [ ] Listed 5+ specific unfair advantages with adjacency zones
  • [ ] Defined constraint envelope (hours, capital, skills, timeline)
  • [ ] Completed 5+ workflow shadowing sessions (90 min each)
  • [ ] Extracted 10+ quantified problem statements
  • [ ] Scored all candidates; kept only ≥18/25
  • [ ] Designed concierge MVP for top 2–3 (input, output, price, timeline)
  • [ ] Secured 3+ paying concierge clients per candidate
  • [ ] Ran 2+ week concierge cycle with tracked metrics
  • [ ] Made automate/pivot/kill decision with documented rationale
  • [ ] Wrote spec from concierge operations log
  • [ ] Shipped thinnest automated slice to existing clients

Closing

Opportunity discovery isn't a creative exercise—it's an observational discipline. The steps above replace guesswork with evidence, and they work whether you're a developer, designer, or operator. If you want a structured workbook that walks through each phase with templates, scoring sheets, and concierge MVP frameworks, start here.

Top comments (0)