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 or blank document, trying to find "a good opportunity." Maybe you want to start a business, build a side project, or identify a new product direction at your company. The advice you've heard is vague: "scratch your own itch," "talk to customers," "look for problems people complain about." But when you actually try to execute, you hit specific walls:

You don't know where to look. Reddit? Twitter? Support tickets? Sales calls? Industry forums? The sheer volume of noise makes it impossible to distinguish signal from complaints that go nowhere.

You can't tell a real problem from a preference. Someone saying "I wish this had dark mode" isn't an opportunity. Someone saying "I lose 4 hours a week manually reconciling these spreadsheets because the export format breaks every time" might be. But in the moment, they sound similar.

You keep finding problems that are already solved. Every time you think you've spotted something, a quick search reveals three funded competitors, an open-source tool, or a feature the incumbent just shipped last month.

You're not sure the problem is worth solving. Even when you find a genuine pain point, you can't answer: Will they pay? Can I reach them? Is the market big enough? Is the switching cost too high?

You're stuck in consumption mode. You read case studies, listen to podcasts, bookmark frameworks—but when it's time to produce your own list of opportunities, the page stays blank.

This isn't a creativity problem. It's a process problem. You're trying to do discovery without a repeatable method.


Why the Usual Fixes Fail

Customer interviews too early. The standard advice—"talk to 20 customers"—assumes you already know who to talk to and what to ask. Without a focused hypothesis, interviews become unfocused chats that confirm what you already believe or leave you with contradictory noise.

Scraping review sites and forums. Mining G2, Capterra, Reddit, or App Store reviews sounds systematic. In practice, you get a list of feature requests and bugs for existing products—not unmet needs. The loudest voices are often power users asking for niche workflows that don't generalize.

Competitor analysis. Mapping competitor features tells you where the market was, not where it's going. It encourages me-too thinking. The best opportunities often look like "bad ideas" to incumbents because they serve a segment the incumbent deliberately ignores.

Trend-chasing. AI, Web3, no-code, creator economy—every cycle produces a wave of founders building for the hype rather than the underlying need. By the time a trend is legible in TechCrunch, the easy opportunities are taken.

Brainstorming sessions. Sticky notes and whiteboards feel productive but produce ideas biased toward what's easy to imagine, not what's valuable to solve. They favor the facilitator's existing mental models.

Waiting for inspiration. The "scratch your own itch" advice works when you have an itch. But most people's personal problems are either too niche (only you have it) or already solved (you just haven't looked hard enough).

What these approaches share: they're either too broad (no filter) or too narrow (premature commitment). You need a middle layer—a systematic way to generate, filter, and validate opportunity hypotheses before you invest serious time.


Step-by-Step Solution

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

Before you collect a single data point, constrain the problem. You cannot search "all problems everywhere."

1.1 Set your constraints explicitly. Write down:

  • Time budget: How many hours per week can you realistically spend for the next 8 weeks?
  • Capital budget: How much can you spend before revenue? (Including opportunity cost of your time.)
  • Distribution advantage: What channels do you already have or can you build? (Email list, SEO domain authority, industry connections, cold email skill, paid ads experience.)
  • Technical capability: What can you build yourself vs. need to hire for?
  • Domain knowledge: What industries, workflows, or user types do you understand deeply?

1.2 Pick 2–3 verticals or user types. Use your constraints to narrow. Examples:

  • "Bootstrapped B2B SaaS for marketing agencies under 50 people" (not "SaaS")
  • "Tools for freelance video editors who bill hourly" (not "creator tools")
  • "Internal tooling for ops teams at Series A logistics companies" (not "B2B software")

1.3 Define your "good opportunity" criteria. Make these measurable. Example:

  • Problem affects ≥500 identifiable businesses/people you can reach
  • They currently pay ≥$200/mo in labor or tools for a workaround
  • You can build an MVP in ≤6 weeks solo
  • Switching cost is low (no data migration, no contract lock-in)
  • You have at least one unfair advantage in distribution or insight

Write these down. They become your filter later.


Phase 2: Systematic Signal Collection (Days 3–14)

Don't "look for ideas." Collect observations in a structured way. Use a spreadsheet with these columns: Source, User Type, Problem Description, Current Workaround, Frequency, Evidence Quote, Your Confidence (1–5).

2.1 Mine your own network (highest signal, lowest volume).

  • Message 10–15 people in your target verticals. Ask: "What's the most tedious, repetitive, or error-prone part of your week? Walk me through the last time it happened."
  • Record exact language. Don't paraphrase.
  • Ask for a screen share or Loom of the workflow if they're willing.

2.2 Shadow support and sales channels (high volume, needs filtering).

  • If you have access: export last 500 support tickets, tag by problem type. Look for "workaround," "manual," "export," "copy-paste," "spreadsheet," "hack."
  • If not: find public communities (Discord, Slack, forums, subreddits) for your vertical. Search for: "how do you handle," "any tool for," "workaround for," "hate that I have to," "wish there was."
  • Save 50–100 raw quotes. Tag each with user type and problem area.

2.3 Analyze job postings and freelance gigs (reveals paid pain).

  • Search LinkedIn, Indeed, Upwork for your vertical. Look for roles titled "Operations," "Admin," "Coordinator," "Specialist" where the description lists manual processes: "manage vendor onboarding spreadsheet," "reconcile daily reports," "manual data entry."
  • These are problems companies already pay to solve poorly. High signal.

2.4 Study adjacent tool ecosystems (reveals integration gaps).

  • Pick 3–5 tools your target users already use (e.g., QuickBooks, HubSpot, Shopify, Notion, Airtable).
  • Browse their app marketplaces, Chrome extensions, Zapier/Make templates. Look for: "sync," "export," "report," "automate," "bulk."
  • Read 1-star reviews of popular integrations. They reveal what the integration doesn't do.

2.5 Track "complaint patterns" not "feature requests."

  • A feature request: "Add dark mode."
  • A complaint pattern: "Every Monday I spend 2 hours copying data from Tool A to Tool B because the CSV export drops custom fields."
  • Only the second is an opportunity. Discard feature requests.

Output goal: 80–150 raw observations in your spreadsheet. Don't filter yet.


Phase 3: Cluster and Score (Days 15–18)

3.1 Cluster by problem, not by tool or feature.
Group observations where the underlying workflow failure is the same, even if described differently.

  • "Export breaks custom fields" + "Can't bulk update tags" + "Manual re-tagging after import" → Cluster: Data fidelity loss during import/export between Tool A and Tool B.

3.2 Score each cluster against your criteria.
Use a simple 1–5 scale for each criterion from Phase 1.3. Example:

Cluster Reachable Users Paying for Workaround Build Time Switching Cost Distribution Fit Total
Data fidelity loss A↔B 5 4 3 5 4 21
Manual vendor

Start here

Top comments (0)