DEV Community

Alan Descure
Alan Descure

Posted on Originally published at smtm-store.pages.dev

Opportunity Discovery: A Practical Guide for Builders Who Keep Building the Wrong Thing

Opportunity Discovery: A Practical Guide for Builders Who Keep Building the Wrong Thing

The Problem in Concrete Terms

You have the skills to build. You've shipped features, maybe whole products. But you keep landing in the same trap: you spend months building something nobody buys.

The pattern looks like this: you notice a vague pain point ("teams struggle with communication"), convince yourself it's a massive market, build a solution, launch to crickets, then rationalize it as "learning." Six months later you're doing it again with a different idea.

The concrete symptoms:

  • Your idea list is either empty or full of "Uber for X" clones
  • You validate by asking friends "would you use this?" and they say yes
  • You can't articulate who specifically has this problem right now and what they're currently doing about it
  • You've never watched a target user work for 30 minutes without interrupting
  • Your "market research" is reading TechCrunch and Reddit threads

This isn't a creativity problem. It's a discovery process problem. You're guessing instead of investigating.


Why the Usual Fixes Fail

1. "Scratch your own itch"
Works if you're a representative user with budget authority. Fails when you're a developer building for developers—you're a weird edge case with high tolerance for jank. Most markets don't look like you.

2. Keyword research / SEO tools
Tells you what people search for, not what they'll pay for. High search volume often means high competition and low urgency. People search for "free project management templates"—they buy "compliance automation for SOC2 audits."

3. Surveying your network
Your network is biased. They like you. They want to be supportive. They'll say "cool idea" to a product they'd never open their wallet for. Politeness is not signal.

4. Building an MVP to "test the market"
An MVP tests execution, not demand. If nobody wants the problem solved, a polished MVP just proves you can build things nobody wants—expensively.

5. Waiting for inspiration
Inspiration is retrospective narrative. Successful founders describe "sudden insight" after months of grinding through customer conversations. The insight came from the work, not before it.


Step-by-Step Solution

Phase 1: Pick a Starting Surface (Days 1–2)

Don't start with "an idea." Start with a constraint that narrows the search space.

Action: Choose one of these three entry points. Not all three. One.

Entry Point Best For Example
Industry you know Domain expertise, existing network You worked in dental offices for 3 years
Buyer you can reach Access to decision-makers Your brother manages 12 HVAC contractors
Workflow you've observed Concrete, visible pain You've watched your spouse do month-end close in Excel

Write your choice at the top of a blank document. This is your search boundary. Everything outside it is out of scope for this cycle.


Phase 2: Map the Workflow (Days 3–7)

You need to see the actual work, not hear about it.

Action: Find 3–5 people who do the target workflow regularly. Not "interested in the space." People who do the thing.

How to get access:

  • Cold DM on LinkedIn: "I'm researching how [role] handles [specific workflow]. Not selling anything. 15 min screen share, I'll buy you coffee."
  • Ask your entry-point contact for introductions: "Who's the most organized / frustrated person you know doing X?"
  • Show up where they are: trade Slack/Discord, industry forums, local meetups

The session structure (30 min max):

  1. Context (5 min): "Walk me through last Tuesday when you did [workflow]."
  2. Observe (15 min): Share screen. Watch. Don't lead. Take timestamped notes: "12:03 — copies data from PDF to Sheet manually," "14:12 — checks email for approval, waits 2 hrs."
  3. Probe (10 min): "At 12:03, what happens if that copy-paste goes wrong?" "What did you do before this tool existed?"

Output: A workflow map per person — steps, tools, handoffs, wait times, workarounds, curse words.


Phase 3: Extract Pain with Evidence (Days 8–10)

Not all pain is equal. You need urgent, expensive, frequent pain.

Action: For each workflow map, mark every step with:

  • Time cost (minutes/hours per occurrence)
  • Frequency (daily, weekly, per client, per month)
  • Error rate (how often it breaks)
  • Current workaround (what they do instead of a tool)
  • Budget signal (have they paid for tools? hired help? built internal scripts?)

Scoring rubric — keep only steps scoring ≥2 on two dimensions:

Dimension 0 1 2 3
Time per occurrence <5 min 5–30 min 30–120 min >2 hrs
Frequency Monthly Weekly Daily Hourly
Error consequence Annoying Rework Lost revenue Compliance/legal risk
Workaround cost None Manual effort Paid tool/VA Custom build
Budget evidence Never paid Free tools only Pays for adjacent tools Has budget line item

Example keeper: "Reconciling Stripe payouts to NetSuite invoices — 45 min/day, daily, 15% error rate causes reconciliation delays, currently using 3-person offshore team ($4.5k/mo)."

Example discard: "Choosing a color for the dashboard — 10 min, weekly, no consequence, no budget."


Phase 4: Verify Willingness to Pay (Days 11–14)

Action: Return to your 3–5 contacts. Present only the high-scoring pain points. No solution. No demo.

Script:

"Based on our conversation, it sounds like [specific step] costs you ~X hours/week and you're currently handling it by [workaround]. If there were a way to reduce that to Y hours without changing [upstream/downstream dependencies], would that be worth a conversation about budget?"

Read the reaction:

  • "Yes, we've actually looked at tools for this" → Strong signal. Ask: "Which ones? Why didn't you buy?"
  • "Maybe, depends on price" → Weak. Ask: "What's the max you'd pay to solve this today?" Silence = no budget.
  • "We've built internal scripts for it" → Strongest. They've already paid with engineering time. Ask to see the script.
  • "Not really a priority" → Discard. Move on.

Goal: 2+ contacts saying "we have budget / we've tried to buy / we've built internally" for the same pain point.


Phase 5: Define the Wedge (Days 15–18)

Don't build the whole solution. Build the smallest slice that delivers measurable value and gets you into the workflow.

Action: Write a wedge spec — one page max:

Problem: [One sentence, user's words]
Current state: [Workflow step + time/cost/error]
Wedge: [Single feature that reduces time/cost/error by ≥50%]
Success metric: [User can measure this themselves in 2 weeks]
Integration points: [What it must connect to — API, CSV, email, nothing]
Excluded: [Everything else — dashboards, settings, auth, reporting]
Enter fullscreen mode Exit fullscreen mode

Example:

Problem: "Reconciling Stripe payouts to NetSuite invoices takes 45 min/day and breaks 15% of the time."
Current state: Manual CSV export → VLOOKUP in Sheets → manual journal entries in NetSuite.
Wedge: Auto-fetch Stripe payouts via API, match to NetSuite invoices by amount+date±1 day, output CSV ready for NetSuite CSV import.
Success metric: Reconciliation time <10 min/day, error rate <2%.
Integration points: Stripe API (read), NetSuite CSV import (write).
Excluded: Dashboard, historical backfill, multi-currency, team roles, Slack alerts.
Enter fullscreen mode Exit fullscreen mode

Phase 6: Pre-Sell the Wedge (Days 19–2

Start here

Top comments (0)