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 you cannot point to a specific problem worth solving. Every idea feels either too small (a feature, not a business), too crowded (ten funded competitors), or too vague ("people need better productivity"). You sit down to brainstorm and produce either a blank page or a list of things you would never actually build. The longer this goes on, the more it feels like a personal failure rather than a process problem.

Concretely, you are stuck in one of three loops:

  1. The consumption loop: Reading "how I found my SaaS idea" posts, scrolling Product Hunt, saving newsletters—never talking to a single potential customer.
  2. The validation theater: Building landing pages, running ads, collecting emails for ideas you don't care about and won't pursue past week two.
  3. The perfection trap: Discarding every idea because you can imagine a reason it might fail, while ignoring that every successful business had fatal flaws at the start.

The output of opportunity discovery is not an idea. It is a problem statement specific enough to build against: "Independent insurance adjusters lose 6+ hours per claim manually reconciling photos with estimate line items." Not "insurance tech is broken."

Why the Usual Fixes Fail

Customer interviews without a hypothesis. You book calls with "founders" or "marketers" and ask "what's your biggest pain point?" They give you polite, vague answers. You learn nothing because you didn't enter with a falsifiable assumption about a specific workflow.

Market sizing too early. You pull TAM numbers from Statista for "global CRM market" and either get excited by billions or discouraged by saturation. Neither tells you whether you can reach these specific customers with this specific solution.

Competitor analysis as idea generation. You study successful companies and try to find "gaps." But the gaps you see from the outside (missing features, ugly UI) are rarely the gaps that matter to users. The real gaps are workflow interruptions you only discover by watching someone work.

Waiting for founder-market fit. You think you need deep domain expertise before starting. In reality, domain expertise comes from doing the discovery work. The founders who "knew the industry" usually learned it by selling into it badly for six months first.

Idea filtering instead of problem collection. You judge each potential problem on "is this a billion-dollar opportunity?" before you've verified the problem exists. You filter out the only tractable starting points because they look small.

Step-by-Step Solution

Phase 1: Choose Your Starting Surface (Days 1-2)

Pick one of these three entry points. Do not combine them.

Option A: Your own workflow friction
List every tool, process, or handoff in your current or recent work that made you think "this is stupid." Be specific: "Jira tickets require manual status updates that Slack already knows" not "project management sucks." Aim for 20+ items. Pick the three that cost you the most time or caused the most errors.

Option B: Adjacent user observation
Identify three people you already know (former colleagues, friends, family) whose work you vaguely understand but don't do yourself. Ask for 45 minutes to watch them work. Not interview—watch. Screen share or sit beside them. Record every time they: switch tools, copy-paste, complain, redo work, or ask a colleague "how do I...?"

Option C: Public complaint mining
Go where your target users complain publicly: Reddit, niche forums, Discord/Slack communities, app store reviews, G2/Capterra negative reviews, Twitter/X searches for "[tool] sucks" or "[workflow] hate." Collect 50+ raw complaints. Cluster them by workflow step, not by tool. Ignore feature requests; keep workflow interruptions.

Decision rule: If you have recent work experience (under 3 years), use Option A. If you have accessible contacts in an interesting industry, use Option B. Otherwise, use Option C. Commit to one for the full process.

Phase 2: Extract Problem Candidates (Days 3-7)

For each friction point from Phase 1, write a Problem Candidate using this template:

[Specific role] spends [time/frequency] doing [manual step] 
because [root cause: missing integration, no automation, bad data, etc.]
This results in [concrete cost: hours lost, errors, delayed delivery, lost revenue].
Current workarounds: [what they do now: spreadsheets, VAs, ignoring it, etc.]
Enter fullscreen mode Exit fullscreen mode

Example:

Independent insurance adjusters spend 45 minutes per claim manually matching photos to line items in Xactimate because photo metadata doesn't import and adjusters must visually match each image. This results in 6+ hours per week per adjuster and frequent supplement requests from carriers due to missing documentation. Current workarounds: naming files manually, printing photos to PDF, hiring admin help at $18/hr.

Write 15-20 Problem Candidates. Do not evaluate them yet. If you cannot fill all five lines for a candidate, discard it—you don't understand the problem well enough.

Phase 3: Rapid Desk Validation (Days 8-12)

For each Problem Candidate, answer these four questions using only public information. Spend max 20 minutes per candidate.

  1. Budget authority: Does this role control budget or influence purchasing? (Search: "[role] software budget," "[role] purchasing decision")
  2. Frequency: Is this daily/weekly (good) or monthly/quarterly (harder)? Confirm via job descriptions, day-in-life posts, forum discussions.
  3. Tool spend: Do they already pay for software? Search "[role] software stack," "[role] tools," "[role] SaaS." If they use zero paid tools, acquisition is harder.
  4. Search behavior: Do they actively search for solutions? Check Google Keyword Planner or Ahrefs for "[problem] software," "[problem] tool," "automate [task]." Zero search volume = they don't frame it as a software problem.

Mark each candidate: Green (yes to 3+), Yellow (2), Red (0-1). Keep only Greens. Expect 3-5 survivors from 15-20 candidates.

Phase 4: Live Validation via Micro-Conversations (Days 13-21)

For each Green candidate, conduct five 15-minute conversations with people matching the role. Not interviews—conversations. Your goal: confirm the problem exists exactly as you wrote it and learn the language they use.

Recruitment message template (LinkedIn, email, warm intro):

"Hi [Name], I'm researching how [role] handle [specific workflow step]. Not selling anything—just trying to understand the current process. Would you spare 15 minutes to walk me through how you do it today? Happy to share what I find."

Conversation script (memorize, don't read):

  1. "Walk me through the last time you did [workflow step]."
  2. "What tools or files did you touch?" (Get specific names)
  3. "Where did it slow down or break?"
  4. "What did you do when that happened?"
  5. "If you could wave a wand and fix one part, what would it be?"
  6. "Who else on your team deals with this?"

Take notes verbatim. Record phrases they use: "reconcile," "match up," "chase down," "clean up." These become your copy later.

Stopping rule: If 3 of 5 people don't recognize the problem as you described it, discard the candidate. If 4+ recognize it but describe it differently, rewrite your Problem Candidate using their language and do 2 more conversations.

Phase 5: Build the Problem Brief (Days 22-24)

For each validated candidate (usually 1-2), write a Problem Brief—the artifact you build against. One page maximum:



PROBLEM BRIEF: [Name]

WHO: [Specific role, team size, industry context]
WORKFLOW: [The 5-7 step process where the problem lives]
BREAKPOINT: [Exact step where friction occurs]
CURRENT COST: [Time/money/errors per unit - use their numbers]
CURRENT WORKAROUND: [What they do now, including tools]
BUYER: [Title who signs off - may differ from user]
TRIGGER EVENTS: [What makes this urgent: audit, growth, new regulation, hiring]
LANGUAGE: [Exact phrases they used - 5-10 quotes]
COMPETITIVE LANDSCAPE: [Tools

**[Start here](https://smtm-store.pages.dev/exp-1794711046/?utm_source=devto&utm_medium=article&utm_campaign=exp-1794711046)**
Enter fullscreen mode Exit fullscreen mode

Top comments (0)