DEV Community

Alan Descure
Alan Descure

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

Opportunity Discovery: A Practical Guide for When You’re Stuck

Opportunity Discovery: A Practical Guide for When You’re Stuck

The Problem in Concrete Terms

You are not "lacking creativity." You are drowning in noise.

You sit down to find a business idea, a product angle, or a content niche. You open Twitter, Reddit, Product Hunt, and three newsletters. Two hours later, you have 40 browser tabs open, a headache, and zero conviction. You see problems everywhere—people complaining about CRM pricing, developers hating their CI/CD pipelines, creators struggling with video editing—but you cannot tell which ones are yours to solve.

The stuckness manifests in three specific ways:

  1. The "Already Exists" Paralysis: Every problem you identify has a funded competitor. You mistake market validation for market saturation and quit before validating your specific angle.
  2. The "Vitamin vs. Painkiller" Guessing Game: You intellectualize the pain level. You convince yourself a "nice-to-have" workflow tool is a burning need because you would use it, ignoring that you are not the target market.
  3. The Source Bias: You only look where you already hang out (Hacker News, Indie Hackers, specific subreddits). You discover opportunities visible to thousands of other builders with your exact same background, guaranteeing high competition for low-differentiation ideas.

You don't need more ideas. You need a filter that separates observable friction from marketable opportunity using assets you already possess.


Why the Usual Fixes Fail

1. "Scratch Your Own Itch"
This works only if your itch is shared by a reachable, paying audience with budget authority. Most developers scratch itches that only other developers have, usually involving tooling that developers expect to be free or open-source. If your itch is "I need a better Notion template," the market size is near zero.

2. Keyword Research / SEO Tools
Ahrefs and Semrush tell you what people search for, not what they will pay to solve. High volume keywords often indicate information-seeking behavior (top of funnel), not purchasing intent (bottom of funnel). You end up building a content site when you wanted a SaaS.

3. "Talk to Customers" (Generic Advice)
Cold DMing strangers on LinkedIn asking "What's your biggest problem?" yields polite lies or vague complaints ("I wish I had more time"). People are bad at articulating root causes and worse at predicting future behavior. Without a structured interview script targeting past behavior, you collect noise, not signal.

4. Trend Surfing (AI, Web3, No-Code)
Building for a trend makes you a feature of the platform, not a business. When the platform shifts (API pricing changes, algorithm updates), your distribution evaporates. Trends are tailwinds for existing businesses; they are quicksand for discovery.


Step-by-Step Solution: The Asset-First Discovery Loop

This process inverts the standard approach. Instead of hunting for a problem then checking if you can solve it, you audit your unfair advantages (assets) first, then hunt only for problems those assets uniquely qualify you to solve.

Step 1: The Asset Inventory (Timebox: 45 Minutes)

Create a spreadsheet with four tabs. Be ruthlessly specific. "Python" is not an asset. "Automating PDF data extraction for logistics invoices using Python/pdfplumber" is an asset.

Tab 1: Technical Capabilities (Hard Skills)

  • Specific stacks you can ship in this week without tutorials.
  • Integrations you have built before (Stripe, Shopify API, HubSpot, AWS Lambda, specific ERPs).
  • Infrastructure you manage confidently (Postgres optimization, Kubernetes, serverless, bare metal).

Tab 2: Domain Access (Network & Knowledge)

  • Industries where you have held a W-2 job for 12+ months.
  • Specific roles you have held (not "marketing"—"demand gen for B2B SaaS $1M-$10M ARR").
  • People you can text right now who are decision-makers (VP, Director, Founder) in specific verticals. Note their title and company size.

Tab 3: Audience & Distribution (Reach)

  • Email list size / open rate (if > 20%).
  • Social followers who engage (replies, DMs), not vanity metrics.
  • Communities you moderate or are a recognized expert in (Discord, Slack, Forums).
  • SEO assets: Domains you own with DR > 30 or existing ranking content.

Tab 4: Capital & Risk Tolerance (Runway)

  • Months of personal runway.
  • Cash available for ads/contractors without stress.
  • Willingness to do sales-led (high touch) vs. product-led (low touch) motion.

Output: A ranked list of your top 3 "Asset Vectors." Example: Vector A: 5 years in dental practice management + network of 12 office managers + ability to build HIPAA-compliant React/Node apps.

Step 2: Targeted Friction Mining (Timebox: 2 Weeks)

Do not "browse." Conduct structured ethnography only inside your Asset Vectors.

Method A: The "Shadow Shift" (Highest Signal)
Ask your 3 strongest contacts (from Tab 2) for 90 minutes to watch them work. Do not interview. Watch.

  • Screen share (remote) or sit behind them (IRL).
  • Record the screen (Loom) and audio.
  • Look for: Workarounds (Excel exports, copy-paste bridges, sticky notes, "I email this to myself"), Complaints to colleagues ("Ugh, I have to re-enter this in System B"), Manual calculations done outside the core software.
  • Rule: You are forbidden from suggesting solutions. Only observe and timestamp the friction.

Method B: The "Switch Interview" (For Past Behavior)
If you can't shadow, interview 5 people who recently switched tools in your vector (last 90 days).

  • Script: "Walk me through the day you decided to cancel [Old Tool]." -> "What did you try first?" -> "What almost stopped you from switching?" -> "What surprised you after switching?"
  • Map the Forces of Progress (Push of current situation, Pull of new solution, Anxiety of change, Habit/Inertia).
  • The opportunity lives in reducing Anxiety or Inertia, not just adding Pull.

Method C: Support Log Archaeology (If you have access)
Export 500+ tickets from a helpdesk (Intercom, Zendesk, Front) for a tool in your vector.

  • Filter for tags: "feature request," "bug," "integration," "workflow."
  • Cluster by job-to-be-done, not feature. "Export to CSV" is a feature. "Reconcile bank statement in Excel because your report doesn't match my bank feed" is a JTBD.
  • Count frequency. High frequency + high manual effort = opportunity.

Step 3: The Viability Filter (Timebox: 2 Hours per Candidate)

Take your top 5 observed frictions. Run each through this Kill/Continue checklist. If it fails any hard constraint, kill it. Move on.

Hard Constraints (Must Pass All):

  1. Budget Authority: Does the person feeling the pain have a corporate card or sign-off power < $500/mo? (If they need CFO approval for $50, kill it unless you love enterprise sales).
  2. Frequency: Does this friction happen daily or weekly? Monthly/Quarterly pains build low-retention tools.
  3. Asset Fit: Can you build the V1 in 4 weeks using only Tab 1 skills? (No "learning Rust" allowed).
  4. Distribution Fit: Can you reach 20 decision-makers this week using Tab 2 or Tab 3? (No "I'll run ads" for discovery phase).
  5. Switching Cost: Is the current solution a spreadsheet, a generic tool (Notion/Airtable/Zapier), or a hated legacy vendor? (If it's a deeply embedded best-in-class tool like Salesforce/Shopify, kill it unless you have a wedge feature).

Soft Constraints (Score 1-5, Need > 12/20):

  • TAM > $10M (Back of napkin: Target Companies × Seat Price).
  • Expansion Revenue Path (Can you upsell seats/usage/modules later?).
  • Founder Motivation (Do you want to talk to these users daily for 3 years?).

Step 4

Start here

Top comments (0)