Opportunity Discovery: A Field Guide for When You’re Stuck Guessing
The Problem in Concrete Terms
You have capacity—time, capital, or a team—but you don’t know where to point it. Every idea feels either too small to matter or too vague to validate. You’re not lacking creativity; you’re lacking a filter. The backlog is full of "maybe" features, "someday" markets, and "interesting" problems that nobody has actually paid to solve.
Concretely, this looks like:
- Spreadsheet paralysis: You have 40 ideas scored on "impact" and "effort" but the top three change every time you re-weight the columns.
- The "Vitamin vs. Painkiller" trap: You can’t tell if a problem is a genuine hair-on-fire emergency or just a mild annoyance people complain about over coffee but won’t budget for.
- Audience opacity: You’re building for "SMBs" or "Developers" or "Marketers"—segments too broad to write a cold email for, let alone a product spec.
- Validation theater: You run surveys asking "Would you use this?" and get polite "Yes" answers that evaporate when you ask for a credit card.
You are stuck because you are treating opportunity discovery as a brainstorming exercise. It is not. It is a research discipline.
Why the Usual Fixes Fail
1. "Scratch your own itch"
This works until you realize you are a niche of one. Your workflow, budget, and tolerance for jank are statistical outliers. Building for yourself usually produces a tool you love that 50 other people find confusing.
2. Keyword research / SEO volume
Search volume tells you what people type into Google, not what they pay for. High volume often means high competition and low intent (informational queries). Low volume often means the problem is too specific to articulate in a search bar—but the buyer is desperate.
3. Competitor feature audits
Copying a competitor’s roadmap means you are always one step behind, fighting for the same saturated budget. You inherit their positioning errors. If they pivoted away from a feature six months ago, you’re building a ghost.
4. "Talk to 100 customers"
Volume of conversations ≠ quality of signal. Most founders ask leading questions ("Wouldn't it be great if...?") and hear confirmation bias. Without a rigid interview script and a definition of "signal" beforehand, 100 chats just produce 100 pages of noise.
5. Waiting for the "Aha!" moment
Insight is not a lightning strike. It is the residue of structured pattern matching across discrete data points. You cannot schedule an epiphany, but you can schedule the work that produces one.
Step-by-Step Solution: The Signal Mining Protocol
This process takes 2–3 weeks of part-time work. It requires zero code, zero ad spend, and zero permission. It produces a ranked list of 3–5 validated problem spaces with identified buyers.
Phase 1: Define the Hunting Grounds (Days 1–2)
Stop looking at "markets." Look at workflows.
- List 10 specific job functions you have access to or understand deeply. Not "Marketing." Email lifecycle marketer at Series B SaaS. Not "Construction." Commercial project manager running $5M–$20M builds.
- Map the "Core Loop" for each. Write the 5–7 repetitive steps they do daily/weekly.
- Example (Email Marketer): Pull segment → Write copy → Build in ESP → QA links/render → Send → Report on clicks → Clean list.
- Flag the "Manual Gaps." Where do they copy-paste between tools? Where do they use a spreadsheet as a bridge? Where do they complain in Slack? These are your entry points. Ignore "feature requests." Hunt for workarounds.
Output: A list of 20–30 specific, observed manual workarounds tied to a named role.
Phase 2: The "Money Map" Filter (Days 3–4)
Filter your workarounds for budget reality.
- Identify the Budget Owner. For each workaround, ask: Who signs the PO for this person’s tools? (Often not the user).
- Find the Line Item. Does this role/department already spend money on software for adjacent steps in the loop?
- Good sign: They pay for an ESP, a design tool, and a deliverability tool. Budget exists.
- Bad sign: They use only free tiers and internal scripts. No budget culture.
- Calculate "Cost of Inaction." Estimate the labor hours the workaround burns per month. Multiply by fully loaded salary. If the workaround costs $2k/mo in wasted engineering time, a $500/mo tool is an easy sell. If it costs $50/mo in junior admin time, it’s a hard sell.
Output: 5–8 workarounds where a Budget Owner exists, a Line Item exists, and Cost of Inaction > $1,000/mo.
Phase 3: The "Unfair Advantage" Interview (Days 5–12)
Do not pitch. Do not demo. Run Problem Discovery Interviews only.
Recruit: 5–8 people per shortlisted workaround. Use LinkedIn Sales Navigator or cold email. Subject line: Research: [Specific Workaround] at [Their Company]. Incentive: $50 gift card or "I’ll share the aggregated anonymized findings."
The Script (Memorize this. Do not improvise):
- Context: "Walk me through the last time you did [Specific Workaround Step]."
- Pain: "What’s the most frustrating part of that specific step?" (Silence. Wait 5 seconds.)
- Current Spend: "What tools or people are currently involved in solving this?"
- Cost: "Roughly how many hours/week does this workaround eat?"
- Magic Wand: "If you could wave a wand and fix only that frustrating part, what changes?" (Note: They often describe a feature, not a product. Dig for the outcome.)
- The Kill Shot: "It sounds like [Problem] costs you ~$X/mo. If a solution existed for $Y/mo that required [Integration Effort], who else needs to sign off?"
- Referral: "Who else in your org hates this workflow as much as you do?"
Rules:
- Record (with permission). Transcribe (Otter/Whisper).
- Zero mention of your solution ideas.
- Score each interview immediately after:
- Urgency (1–5): Are they actively looking now?
- Budget Access (1–5): Can they name the approver?
- Workaround Severity (1–5): Is it duct tape or a missing limb?
Output: A spreadsheet of interviews tagged with scores. Keep only problems scoring ≥12/15 across 3+ interviewees at the same company or 5+ interviewees across companies.
Phase 4: The "Concierge" Validation (Days 13–18)
You do not build software yet. You sell the outcome manually.
- Define the Service. "I will [perform the specific workaround output] for you for 2 weeks for $500." (Price it at 10x your hourly rate to test value, not affordability).
- Pitch the 3 Best Prospects. Email the interviewees who scored highest: "Based on our chat, I’m running a private beta for a done-for-you version of [Workflow]. 2 weeks, $500, you get [Specific Deliverable]. Interested?"
- Execute. Do the work manually. Use Zapier, Sheets, APIs, VA labor, your own hands. Deliver the result.
- Measure:
- Did they pay? (Invoice sent → Paid = Hard Yes).
- Did they use the output? (Did they forward the report? Import the CSV?).
- Did they ask for week 3?
Output: 0–3 paying concierge clients.
- 0 Payers: Kill the problem. Return to Phase 1 list.
- 1–2 Payers: You have a problem worth solving. Proceed to spec.
- 3+ Payers: You have a business. Start automating the highest-leverage piece.
Phase 5: Spec the "Wedge" (Days 19–21)
Do not build the "platform." Build the wedge—the single feature that replaces the highest-value manual

