Frameworks

Methodology

Decision frameworks for evaluating ideas, building products, and knowing when to kill or continue. These aren't rules—they're defaults that can be overridden with good reason.

Starting

Idea Evaluation

What makes an idea worth pursuing? Run through these criteria before committing time.

✓

Clear, specific pain point

Not 'would be nice' but 'this is broken and I need it fixed'

Test: Can you name 3 people who have this problem and would pay today?

✓

AI changes what's possible

Not just automation of existing process, but genuinely new capability

Test: Could this exist without recent AI advances? If yes, why hasn't it?

✓

Path to revenue without sales team

Self-serve, product-led, or clear inbound channel

Test: How does someone find this, try it, and pay—all without talking to anyone?

✓

Solo-buildable and testable

One person can ship v1 and get signal within 2-4 weeks

Test: What's the smallest thing you can ship that proves/disproves the thesis?

Red Flags

✗'Lots of people might want this' (too vague to validate)
✗Requires marketplace dynamics or network effects to be valuable
✗Needs enterprise sales motion to close deals
✗Can't articulate who pays and why in one sentence
✗Requires significant upfront investment before any signal
✗Success depends on partnerships or platform access

Building

Project Phases

Every project moves through these phases. Each has a clear goal and exit criteria.

Evaluating

1-3 days

Goal: Decide whether to start

• Research existing solutions
• Identify potential early users
• Sketch minimum viable scope
• Estimate build effort

Exit: Clear thesis + scope doc, or explicit pass

In Progress

1-2 weeks

Goal: Ship something testable

• Build core functionality
• Deploy to production URL
• Instrument basic analytics
• Prepare distribution assets

Exit: Live product real users can try

Validating

2-4 weeks

Goal: Get signal on thesis

• Drive traffic through chosen channels
• Monitor usage and engagement
• Collect qualitative feedback
• Test pricing if applicable

Exit: Clear continue or kill decision with reasoning

Shipped

Ongoing

Goal: Sustain and grow

• Iterate based on feedback
• Expand distribution
• Add features that retain users
• Consider next growth lever

Exit: Revenue or strong growth trajectory

Deciding

Kill / Continue Framework

The hardest part isn't building—it's knowing when to stop. Use these signals.

Continue If

Someone paidStrongest

Even $1 from a stranger is signal. Friends and family don't count.

Clear path to paymentStrong

Users engaged + specific blockers identified + path to remove blockers

Learning valueModerate

Even if product fails, skills/patterns transfer to next attempt

Kill If

No usage despite distribution

If you can't get anyone to try it, the problem might not be acute enough

Usage but no payment intent

Vitamin not painkiller—nice to have but not worth paying for

Core assumption invalidated

The technical approach doesn't work, or the market doesn't exist

Better opportunity identified

Opportunity cost is real—pivot to higher-signal idea

"Never kill just because it's hard. Kill because the thesis is wrong."

Validating

Distribution Channels

Test channels in parallel with small bets. Double down on what works, drop what doesn't.

Channel	Type	Effort	Notes
Product Hunt	Launch	Medium	Good for initial burst. One-shot—make it count.
Indie Hackers	Community	Low	Build in public. Good for feedback, moderate for traffic.
Twitter/X	Social	High (ongoing)	Requires consistent presence. Better if you already have audience.
Cold outreach	Direct	High	Low conversion but high signal. Good for B2B validation.
Content/SEO	Inbound	High (delayed)	Compounds over time. Not for quick validation.
Marketplaces	Platform	Medium	App stores, plugin directories. Built-in audience if you fit.

Expectations

Time & Cost Benchmarks

Rough targets for how long things should take and cost. If you're way over, question assumptions.

Evaluation

1-3 days

Build (v1)

1-2 weeks

$50-200 (hosting, APIs)

Validation

2-4 weeks

$100-500 (ads, tools)

Principles

→Speed matters more than perfection—ship fast, iterate faster
→AI development is 10-20x faster than traditional estimates
→Most of the 'work' is deciding what to build, not building it
→If you're spending more than $500 before first revenue, question the thesis