Opportunity Discovery: A Practical Guide for When You're Stuck
The Problem in Concrete Terms
You have the capacity to build something. You have skills, maybe some savings, definitely the desire. But every idea you generate feels either obvious and crowded, or vague and untestable. You sit down to "find an opportunity" and end up doom-scrolling Product Hunt, reading about AI wrappers raising millions, or convincing yourself that the only real problems left require a PhD and a research lab.
The stuckness looks like this:
- You've listed 20 ideas in Notion. None survive a 10-minute pressure test.
- You can't tell if a problem is "real" or just something you think people have.
- You've talked to three potential users. They were polite. You learned nothing.
- You keep waiting for a "pull" signal—someone begging for a solution—but silence feels like evidence of no opportunity.
- You oscillate between "everything's taken" and "I could build that in a weekend" without ever committing.
This isn't a creativity deficit. It's a process deficit. You're trying to have ideas instead of deriving them.
Why the Usual Fixes Fail
1. "Scratch your own itch"
Works if your itch is shared by thousands of people with budget. Fails when your itch is idiosyncratic, too small, or already solved by a tool you haven't heard of. Most developers' itches are terrible business opportunities.
2. "Talk to customers"
Generic advice. If you don't know who to talk to or what to ask, you get polite nods and zero signal. "Would you use a tool that...?" produces lies. People want to be helpful. They'll say yes to anything that sounds reasonable.
3. "Look for complaints on Reddit/Twitter"
You find noise. People complain about everything. The loudest complaints are often low-value, high-effort problems (e.g., "I hate filing taxes") where the barrier isn't software—it's regulation, trust, or switching costs. You waste weeks validating problems that can't be solved with code.
4. "Build an MVP and launch"
You build the wrong thing fast. The market doesn't reward speed on the wrong vector. You get silence, interpret it as "need more features," and build further in the wrong direction.
5. "Follow trends (AI, Web3, no-code)"
Trend-chasing puts you in the most crowded arena with the least differentiation. By the time a trend is legible on Twitter, the alpha is gone. You're competing with funded teams who started 18 months ago.
Step-by-Step Solution
Step 1: Define Your Constraints First (30 minutes)
Before looking outward, write down your non-negotiables. This filters 90% of opportunities instantly.
Answer in writing:
- Time: How many hours/week can you consistently invest for 6 months? (Be honest. "20" usually means 5.)
- Capital: How much can you spend without stress? $0? $5k? $50k?
- Skills: What can you actually build well right now—not what you could learn.
- Distribution: Do you have any audience, network, or channel access? (Zero is a valid answer.)
- Revenue floor: What's the minimum monthly revenue that makes this worth continuing? $500? $5k?
- Exit intent: Are you building to sell, to run indefinitely, or to learn?
Example constraint set:
10 hrs/week, $2k capital, strong backend/weak frontend, zero audience, $2k/mo floor, run indefinitely.
Keep this visible. Every idea gets tested against it.
Step 2: Map Adjacent Problems You Can Actually See (2 hours)
Don't hunt for problems. Map the ones you're already near.
List every domain where you have:
- Professional experience (current/previous jobs, freelance clients)
- Personal obsession (hobbies you've spent 100+ hours on)
- Second-hand access (spouse's industry, friend's business, parent's workflow)
- Past failed attempts (you already know the landscape)
For each domain, write specific workflows—not "marketing" but "how a 12-person B2B agency onboards clients." Not "fitness" but "how powerlifters track accessory volume across mesocycles."
Output: 15-30 specific workflows. Each should be describable in one sentence: "Who does what, with which tools, and where it breaks."
Step 3: Run the "Manual First" Filter (1 hour per workflow)
For each workflow, answer without researching:
- Who exactly does this? (Title, company size, seniority)
- What tool(s) do they use today? (Name them. "Excel," "Notion," "Jira + custom scripts.")
- Where exactly does it break? (Step in the workflow, frequency, consequence)
- How do they currently cope? (Workaround, hire, ignore, pay for)
- What triggers them to seek a new solution? (New hire, growth, regulation, pain threshold)
- Can I reach 5 of these people this week? (Yes/No/How)
Discard any workflow where:
- You can't name the current tools
- The break point is vague ("it's slow")
- The cope is "they ignore it" (no urgency)
- You can't reach 5 people this week
Keep 3-5 workflows. These are your candidate opportunities.
Step 4: Conduct 5 "Workflow Interviews" Per Candidate (1-2 weeks)
Do not pitch. Do not ask "would you use." Reconstruct their actual behavior.
Recruit:
- Cold DM/email: "I'm researching how [title] handle [specific workflow]. Not selling anything. 15 min, I'll send you a $20 gift card."
- Use your network, LinkedIn, relevant Slack/Discord, industry forums.
- Target people who did the workflow yesterday.
Interview script (record, transcribe later):
- "Walk me through the last time you did [workflow]." (Let them narrate. Don't interrupt.)
- "What tools did you use at each step?"
- "Where did it get annoying or risky?"
- "What did you do about that annoyance?" (Key: did they pay, build, hire, tolerate?)
- "Tell me about the last time you looked for a better way." (What triggered it? What did they evaluate? Why did they stay/leave?)
- "If magic wand—what changes?" (Listen for workflow changes, not feature requests.)
After each interview, write a 3-bullet summary:
- The exact break point (quote them)
- The current cope + cost (time/money/risk)
- The trigger that would make them switch
Stop when: You can predict the next interviewee's answers. Usually 5-8 per workflow.
Step 5: Score and Select (1 hour)
Build a simple spreadsheet. One row per workflow. Columns:
| Workflow | Break Point Clarity (1-5) | Urgency (1-5) | Willingness to Pay Evidence (1-5) | Reachability (1-5) | Fits Constraints (Y/N) | Total |
|---|
Scoring rules:
- Break Point Clarity: 5 = they described the exact step + consequence in same words across interviews. 1 = vague "it's frustrating."
- Urgency: 5 = they've already paid for a partial solution or built internal tooling. 1 = "we live with it."
- Willingness to Pay: 5 = named a dollar figure or showed invoice. 1 = "maybe if it's cheap."
- Reachability: 5 = you have a list of 50+ names and a warm intro path. 1 = cold outreach only.
- Fits Constraints: Binary. If building the obvious solution exceeds your time/skill/capital, it's N.
Pick the highest score that fits constraints. Not the "best" idea—the one you can actually execute and validate.
Step 6: Define the "Slice One" Scope (30 minutes)
Don't build the solution. Build the smallest slice that proves the break point is real and your approach works.
Write a one-pager:
- Target user: Exact title + context (e.g., "Founder-led B2B agencies 5-15 people")
- Break point: One sentence from their words
- Current cope + cost: What they do now + what it costs
Top comments (0)