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 to matter, too crowded to enter, or too vague to act on. You scroll through Product Hunt, Indie Hackers, and Twitter threads hoping for a spark. You save articles about "profitable micro-SaaS niches" and never build anything. You've run customer interviews that yielded polite nods but no urgency. You've built two MVPs that nobody used. The common thread: you're guessing at demand instead of locating it.
Opportunity discovery isn't "ideation." It's the systematic work of finding a specific group of people with a specific, expensive, recurring problem they're already paying to solve badly. When you're stuck, it's usually because you're starting from your capabilities and working outward, rather than starting from their pain and working inward.
Why the Usual Fixes Fail
1. "Scratch your own itch" works only if your itch is shared by enough people who pay for solutions. Most personal itches are either too niche or already solved.
2. Keyword research and SEO tools show search volume, not willingness to pay. High volume often means low intent or high competition. Low volume often means no market.
3. "Uber for X" / "Stripe for Y" analogies skip the part where you verify the underlying workflow exists and is painful. Most analogies map to markets that don't exist.
4. Asking people "would you use this?" generates false positives. People are polite, hypotheticals are free, and stated intent correlates poorly with actual purchase behavior.
5. Building an audience first delays the only validation that matters: someone handing you money for a solution to a problem they have today.
6. Trend-chasing (AI, Web3, no-code, etc.) attaches you to a technology cycle, not a customer problem. When the hype fades, the problem remains—or never existed.
Step-by-Step Solution
Phase 1: Define Your Search Boundaries (Days 1–2)
Before hunting, constrain the space. You cannot search "all problems." Write down:
- Budget: How much revenue do you need this to generate in month 12? ($3k? $30k? $300k?)
- Time: How many hours/week can you commit for 6 months without revenue?
- Domain access: Which industries, roles, or communities do you already have credible access to? (Not "I know a guy." You need: "I worked in this industry for 3 years" or "I moderate a 2k-member Slack for X.")
- Technical ceiling: What can you build alone in 4–6 weeks? Be honest.
- Dealbreakers: What will you not do? (No cold outreach? No regulated industries? No hardware? No B2C?)
Output: A one-page "search brief" you can reference when evaluating leads. If an opportunity violates a boundary, discard it without guilt.
Phase 2: Map Workflows, Not Markets (Days 3–10)
Pick one industry/role from your access list. Your goal: document the actual work people do, step by step, and where it breaks.
Method A: Shadowing (best)
- Ask 3–5 contacts if you can sit in on 60–90 minutes of their work. "I'm researching how [role] handles [process]. Can I watch you work for an hour? I'll buy lunch."
- Take timestamped notes: What tool? What input? What output? Where do they pause, swear, switch tabs, copy-paste, export to Excel, complain to a colleague?
- Do not pitch. Do not suggest fixes. Just observe.
Method B: Structured Interviews (if shadowing isn't possible)
- Recruit 8–12 people via your network. Pay $50–100 for 45 minutes.
- Script:
- "Walk me through the last time you did [specific process] from start to finish."
- "What tools did you use at each step?"
- "Where did it go wrong, take too long, or require a workaround?"
- "What did you do when it broke?"
- "Have you paid for a tool to fix this? Which one? What's missing?"
- "If you could wave a wand and change one thing about this workflow, what would it be?"
- Record (with permission). Transcribe. Highlight every mention of: manual work, spreadsheets, "I wish," "we pay for," "it breaks when," "nobody owns this."
Output: A workflow map for one core process, annotated with friction points, current spend, and workarounds.
Phase 3: Score Each Friction Point (Days 11–13)
For every friction point identified, score 1–5 on:
| Criterion | 1 | 3 | 5 |
|---|---|---|---|
| Frequency | Monthly or less | Weekly | Daily or multiple/day |
| Cost of failure | Annoying | Delays output, costs hours | Loses revenue, risks compliance, loses customers |
| Current spend | $0, free tools only | Pays for partial tools | Pays $100+/mo/user or hires contractors |
| Workaround complexity | None | Spreadsheet + manual steps | Custom scripts, dedicated admin, outsourced |
| Decision maker accessibility | Unknown, gatekept | Reachable via warm intro | You know them directly |
| Technical feasibility | Requires team + 6 months | Buildable solo in 4–8 weeks | Buildable solo in 2–3 weeks |
Keep only items scoring ≥20/30. Discard the rest. You should have 1–3 candidates.
Phase 4: Verify Willingness to Pay (Days 14–21)
For each remaining candidate, run a paid validation test. Not a survey. Not a waitlist. A test where money or committed time changes hands.
Option A: Concierge / Manual Service
- Offer to do the work manually for 3–5 target users for 2–4 weeks.
- Price: 50–75% of the tool cost you'd eventually charge.
- Deliver the output they currently struggle to produce.
- If they pay and renew: you have a problem worth automating.
- If they ghost: the pain isn't urgent enough.
Option B: Pre-sell a "Founding User" Spot
- Create a one-pager: problem, proposed solution, timeline (8–12 weeks), price (lifetime or 12-month), cap (5–10 spots).
- Send to 20–30 qualified contacts (from interviews + referrals).
- Ask for a 50% deposit via Stripe link to reserve.
- If 2+ pay: build. If 0–1 pay after 2 weeks of follow-up: kill it.
Option C: Paid Pilot with Existing Tooling
- If a partial tool exists (Zapier, Airtable, Notion, Retool), build a custom version for one company in 1–2 weeks.
- Charge a setup fee + monthly retainer.
- Document every edge case they hit. That's your product spec.
Output: At least one paying pilot customer, or a documented "no" that frees you to move to the next candidate.
Phase 5: Build the Smallest Thing That Removes the Workaround (Days 22–60)
Do not build "the product." Build the automation of the highest-scoring workaround.
- If they copy-paste from Tool A → Excel → Tool B: build the sync.
- If they manually classify tickets: build the classifier.
- If they chase approvals in Slack: build the routing + reminders.
- If they reconcile spreadsheets monthly: build the reconciliation report.
Constraints:
- Single user role, single workflow, single integration.
- No auth system (use magic links or shared accounts).
- No settings pages (hardcode defaults, expose via config file).
- No dashboard (email/Slack the output).
- Deploy to a $5 VPS or Railway/Render free tier.
- Bill via Stripe Payment Link (no subscription logic yet).
Goal: The pilot customer stops using the workaround. Measure: time saved per occurrence × frequency × their hourly cost. If >$200/mo value delivered, you have a business.
Phase 6: Expand Outward from the Wedge (Month 3+)
Only after the wedge works:
Top comments (0)