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 and crowded, or obscure and unvalidated. You've read about "scratching your own itch," but your itches are either too niche (a better keyboard layout for left-handed Dvorak users) or too saturated (another project management tool). You've tried listing problems you encounter daily, but they're either trivial (forgot to buy milk) or structural (healthcare costs) with no clear entry point.

The stuckness manifests in specific ways:

  • Idea paralysis: You have a Notion page with 47 half-baked concepts, none developed past the first paragraph
  • Validation avoidance: You keep "researching" instead of talking to anyone because you're not sure what to ask
  • Market confusion: You can't tell if a market is "too competitive" or "proven demand" — both look like crowded spaces
  • Scope creep: Every idea balloons into a platform before you've validated a single feature
  • False starts: You've built three MVPs this year. None have paying users. All are technically functional.

The core issue isn't creativity—it's that you lack a repeatable process for moving from vague interest to testable hypothesis to evidence-backed opportunity.


Why the Usual Fixes Fail

1. "Scratch your own itch"
Works if your itch is shared by thousands of people with budget and urgency. Fails when your problems are idiosyncratic, or when you're solving a problem you had three years ago but no longer feel acutely. You lose the visceral signal that guides product decisions.

2. "Look for complaints on Reddit/Twitter/forums"
You find real pain, but no context on willingness to pay, budget authority, or switching costs. A thousand complaints about Jira doesn't mean teams will buy your alternative—it means they're stuck with Jira. Complaints ≠ demand.

3. "Follow the money / look at successful companies and find gaps"
Survivorship bias. You see the winners, not the graveyard of similar attempts. The gap you spot might be a trap (regulated, high switching costs, requires network effects you can't bootstrap).

4. "Customer development interviews" (as typically practiced)
You ask "would you use this?" and get polite lies. You ask "what's your biggest problem?" and get vague complaints unactionable for product design. Without a structured hypothesis, interviews become therapy sessions for the interviewee.

5. "Build an audience first"
Takes 12-24 months of consistent content. You need signal now. Also reverses the logic: you're building distribution for a product you haven't validated exists.

6. Trend-chasing (AI, Web3, no-code, etc.)
By the time a trend is legible enough for you to spot it, the low-hanging fruit is gone. You're competing with funded teams who started 18 months ago.


Step-by-Step Solution

This process takes 3-4 weeks if you treat it as a part-time job (10-15 hrs/week). It produces a ranked shortlist of 2-3 opportunities with enough evidence to justify building an MVP—or killing the idea before you write code.

Phase 1: Define Your Constraints (Week 1, 2-3 hours)

Before generating ideas, constrain the search space. Write explicit answers to:

Financial constraints

  • Minimum monthly revenue to make this worth your time: $____
  • Runway without revenue: ____ months
  • Maximum upfront capital you'll invest: $____

Operational constraints

  • Hours/week you can commit: ____
  • Skills you can apply without learning curves: ____
  • Skills you're willing to learn (max 1 major new skill): ____
  • Do you want to sell B2B, B2C, B2B2C, or marketplace? ____
  • Geographic constraints (timezone, language, regulation): ____

Strategic constraints

  • Exit horizon: lifestyle business / acquisition target / VC-scale / unsure
  • Distribution channels you have or can build: SEO, cold outbound, paid ads, partnerships, marketplace, content, other: ____
  • Competitive moats you can realistically build: switching costs, network effects, brand, data, IP, none: ____

Non-negotiables

  • Problems you refuse to work on (ethics, boredom, regulation): ____
  • Customer types you refuse to serve: ____

Output: A one-page "Opportunity Filter" document. Every idea must pass this filter before you spend a minute validating it.


Phase 2: Generate Hypotheses, Not Ideas (Week 1-2, 6-8 hours)

Stop brainstorming products. Start formulating testable hypotheses in this format:

Hypothesis: [Specific customer segment] has [specific, urgent, expensive problem] and currently solves it with [inadequate solution]. They would pay [price range] for [specific value proposition] delivered via [distribution channel].

Sources for hypotheses (prioritized):

  1. Your recent paid work (last 18 months): What did clients complain about that wasn't in scope? What did you hack together internally? What did you see them pay consultants for repeatedly?
  2. Adjacent roles: If you're a dev, talk to designers, PMs, sales, support, ops. What tools do they hate? What processes run on spreadsheets?
  3. Vendor switching events: Companies that recently migrated CRM, ERP, HRIS, analytics. The migration pain is the opportunity—either a tool that eases migration, or a better alternative for the next switch.
  4. Regulatory/compliance changes: New requirements (SOC2, GDPR, HIPAA, accessibility, ESG) create forced demand. Look at enforcement actions, not just regulations.
  5. Platform dependency risks: Businesses built on Shopify, Amazon, Apple, Google, Salesforce, AWS. What happens when the platform changes pricing, API, or policy? That fear is budget.
  6. High-churn categories: Look at G2/Capterra categories with >20% annual churn. High churn = unsolved problem = opportunity for a focused alternative.

Generation method: Set a timer for 90 minutes. Write 20 hypotheses. Don't filter. Use the template. If you can't fill all brackets, the hypothesis is incomplete—note what's missing.

Output: 20 written hypotheses in a spreadsheet with columns: Segment, Problem, Current Solution, Price Hypothesis, Channel Hypothesis, Source, Missing Info.


Phase 3: Rapid Disqualification (Week 2, 4-5 hours)

Apply your Opportunity Filter from Phase 1. Kill ruthlessly. Aim to eliminate 12-15 hypotheses in 30 minutes.

Then for the remaining 5-8, fill missing info with desk research only (no calls yet):

  • TAM proxy: LinkedIn Sales Navigator count of target titles in target geography × estimated budget. Rough order of magnitude.
  • Competitive density: Search "[problem] software" + "[problem] tool" + "[problem] alternative". Count unique vendors on page 1-2. >15 = crowded. <5 = possibly no market.
  • Search volume: Ahrefs/SEMrush/Ubersuggest for 10-15 problem-related keywords. Monthly volume <500 = weak pull. >10,000 = expensive clicks.
  • Hiring signals: Indeed/LinkedIn Jobs for "[problem] specialist" or "[tool] admin". Growing headcount = budget exists.
  • Review mining: Read 20 negative reviews of top 3 competitors on G2/Capterra. Extract specific complaints. Do they match your hypothesis?

Output: Scored shortlist of 3-4 hypotheses with completed data. Score each 1-5 on: Market Pull, Budget Clarity, Access to Decision-Makers, Competitive Gap, Founder Fit. Keep top 2-3.


Phase 4: Problem Validation Interviews (Week 3-4, 8-12 hours)

Goal: Confirm the problem exists, is urgent, has budget, and your hypothesized segment actually experiences it in the way you think.

Recruit 8-12 people per hypothesis (16-24 total). Sources:

  • LinkedIn: Search target title + company size. Connect with note: "Researching [problem] for [segment]. Not selling. 15 min call, $50 gift card."
  • Cold email: 50-100 per hypothesis. Short, specific, transparent.

- Communities: Slack, Discord, Discourse, Reddit (DM, don't post).

Start here

Top comments (0)