Opportunity Discovery: A Field Guide for When You’re Stuck Guessing
The Problem in Concrete Terms
You have capacity—time, capital, or a team sitting idle—but every direction looks either crowded, vague, or like a distraction. You aren't lacking ideas; you are drowning in unvalidated ones. You have a Notion page with 40 "potential markets," a spreadsheet of TAM calculations you don't trust, and a browser history full of competitor pricing pages.
The stuckness manifests in three specific ways:
- The "Shiny Object" Loop: You spend two days deep-diving a niche (e.g., "AI for dental scheduling"), build a mental model, get excited, then find a well-funded competitor or realize the sales cycle is 18 months. You discard it and start over.
- The "Internal Bias" Trap: You evaluate opportunities based on what you find technically interesting or easy to build, ignoring whether the buyer has budget authority or acute pain right now.
- The "Data Paralysis" Stall: You try to "analyze" your way to certainty. You pull Semrush data, G2 reviews, and LinkedIn headcount growth. The spreadsheet grows, but the decision doesn't get made. You are using research as a substitute for risk.
The core issue: You are treating discovery as a research problem. It is a sales problem. You cannot discover an opportunity by reading about it. You discover it by finding someone willing to pay for the risk to disappear.
Why the Usual Fixes Fail
1. "Scratch Your Own Itch"
This works for side projects, not businesses. Your pain is idiosyncratic. You are not a representative sample. Building for yourself usually yields a tool with a market of one (you) or a crowded category where "better UX" is the only differentiator—a losing battle against incumbents with distribution.
2. Keyword Research / SEO Gap Analysis
This tells you what people search for, not what they buy. High volume, low competition keywords often indicate informational intent (blogs, tutorials) or markets with no budget (hobbyists, students). It optimizes for traffic, not revenue.
3. Trend Reports & Market Research (Gartner, Forrester, CB Insights)
These are lagging indicators. By the time a trend appears in a $5,000 report, the early adopters have already locked in vendors. You are buying a map of where the gold was.
4. "Build an MVP and Launch on Product Hunt / Hacker News"
This validates interest from builders, not demand from buyers. You get signups from people who like launching products, not people who have the specific, expensive problem you solve. Churn follows immediately.
5. Cold Outreach to "Validate"
Sending 500 DMs asking "Would you use a tool that...?" gets you polite lies. People are nice. They say "Yes, sounds cool" to end the conversation. This is not validation; it is noise.
Step-by-Step Solution: The "Buyer-First" Discovery Loop
This process takes 2–3 weeks. It requires zero code, zero ad spend, and zero "networking events." It requires talking to strangers who spend money.
Phase 1: Define Your "Unfair Advantage" Constraints (Day 1)
Before looking outward, audit inward. You need boundaries to filter the infinite opportunity space. Write down honest answers to these four questions. Do not aspire; describe reality.
- Technical/Domain Moat: What specific, hard-to-replicate knowledge, IP, data access, or regulatory navigation do you/your team possess right now? (e.g., "We know HIPAA audit logging inside out," "We have a proprietary parser for legacy EDI 835 files," "My co-founder ran a 200-person roofing supply chain.")
- Channel Access: Where can you get a meeting this week without a cold email? (e.g., "My old boss is VP Ops at a logistics co," "I moderate a Slack for 2k indie insurance adjusters," "I speak at the annual Dental Lab conference.")
- Budget Reality: What is the minimum Annual Contract Value (ACV) you need to make the math work for your runway/team size? ($5k? $50k? $200k?)
- Time-to-Revenue Ceiling: How many months until you need cash in bank? (3? 6? 12?)
Output: A one-page "Filter Card." Every opportunity must pass this filter. If it requires knowledge you don't have, channels you can't open, ACV too low, or sales cycle too long—discard it immediately.
Phase 2: Build the "Suspect List" (Days 2–3)
Do not look for "markets." Look for buyers with the specific problem your Filter Card unlocks.
Action: Create a spreadsheet with 50 rows. Columns: Company Name, Specific Role (Title), Name, LinkedIn URL, Evidence of Pain, Channel In.
How to fill 50 rows in 4 hours without a database:
- LinkedIn Sales Nav (Free Trial/Boolean): Search:
("Head of Operations" OR "VP Operations" OR "Director of Supply Chain") AND ("Logistics" OR "Freight Forwarding") AND ("Series A" OR "Series B" OR "51-200 employees"). Filter by "Posted in last 30 days" (active buyers). - Job Boards (Indeed, LinkedIn Jobs, Niche Boards): Search for roles hiring for the problem you solve. "Hiring: Manual Data Entry Specialist," "Hiring: Prior Authorization Coordinator." The company has budget (they're paying salary) and proven pain (they're hiring humans to patch process).
- Conference Attendee Lists / Sponsor Pages: Download the PDF from last year's industry event. Sponsors have budget. Attendees have intent.
- Clay / Apollo / BuiltWith (Free Tiers): Enrich domains for tech stack signals. (e.g., "Uses NetSuite + Shopify + ShipStation" = complex ops stack).
Crucial Filter: Every row must have a filled Evidence of Pain column.
- Bad: "Big company."
- Good: "Job post: 'Automate 500 manual invoice reconciliations/month'."
- Good: "CEO quoted in FreightWaves: 'Our biggest bottleneck is carrier onboarding compliance.'"
- Good: "Uses 3 different TMS platforms (per BuiltWith) = integration hell."
Phase 3: The "Problem Interview" (Days 4–14)
Goal: Not to pitch. Goal: Map the current workflow and find the "bleeding neck" (urgent, expensive, frequent problem).
The Script (Memorize the flow, not the words):
- Context: "I'm researching how [Specific Role] at [Company Size/Type] handle [Specific Workflow implied by Evidence of Pain]. Not selling anything. 15 mins?"
- Current State: "Walk me through the last time you did [Workflow]. Step by step. What tools? Who touches it?"
- The Friction: "Where does it break? What’s the manual hack? How often?"
- The Cost: "When it breaks, what happens? Lost revenue? Compliance fine? Overtime? Churned client? Put a dollar number on the last incident."
- Current Solutions: "What have you tried? (Software, consultants, hiring, scripts). Why did it fail/stay?"
- Buying Motion: "If a solution existed that solved just [Specific Breakpoint], who signs the PO? What’s the process? Security review? Legal?"
- The "Magic Wand" Close: "If you could wave a wand and fix one thing in this workflow tomorrow, what is it? Why that one?"
Execution Rules:
- 5 calls minimum before writing a line of code/spec.
- Record (with permission) + Transcribe (Otter/Fathom).
- Score immediately after call:
-
Pain Level (1-5):Is it a "nice to have" (1-2) or "get fired if not fixed" (4-5)? -
Budget Evidence (Y/N):Did they mention current spend (tools, headcount, agencies)? -
Access (Y/N):Can you reach the decision maker? -
Fit (Y/N):Does it match your Filter Card?
-
Top comments (0)