The Kill/Greenlight Ritual: How We Stay Ruthless About Portfolio Quality

Most founders can't kill ideas because of sunk costs. We make tough decisions systematically, not emotionally.

I'll never forget the moment I finally killed my "revolutionary AI-powered meal planner."

I'd spent 3 months on it. Built a working prototype. Talked to 50 users. Even got 200 waitlist signups. The validation looked decent—not amazing, but decent.

But when I looked at the numbers honestly:

Portfolio Score: 22/40 (barely above kill threshold)
Opportunity Score: 7.2/10 (below 8.0 proceed threshold)
Validation Results: 200 signups (threshold: 500), 4% prepayment conversion (threshold: 10%)

Every rational signal said "kill this." But I'd invested 3 months. I'd told friends about it. I'd built features I was proud of. The sunk cost fallacy screamed "just give it one more week."

I killed it anyway.

Two weeks later, I started working on the habit tracker (scored 32/40, Opportunity 8.5/10). It hit every validation threshold in 4 weeks. Now it's in production.

The meal planner would have consumed 6 more months and failed anyway. The habit tracker succeeded because I started it instead.

That's the power of ruthless portfolio management: saying no to mediocre ideas so you can say yes to exceptional ones.

In this article, I'll show you exactly how we run kill/greenlight reviews, the criteria we use to make tough calls, how we resist sunk-cost fallacy, and how the Hub makes these decisions visible and systematic.

The Problem: Founders Can't Kill Ideas

Let's start with why most founders struggle to let go.

The Emotional Traps

1. Sunk Cost Fallacy

"I've already spent 3 months on this. I can't quit now."

Why it fails: Past time is gone. The question isn't "was it worth 3 months?" but "is it worth the next 3 months?"

Example: You've built 60% of an MVP. The validation is weak. Do you:

A) Finish the last 40% (3 more months) and launch to mediocre results
B) Kill it now and start a better idea (3 months → potential success)

Sunk costs push you toward A. Rational expected value pushes you toward B.

2. Identity Attachment

"This is my idea. I came up with it. It defines me."

Why it fails: Your identity shouldn't be tied to one idea. You're a builder, not "the person who builds meal planners."

Example: You tell everyone about your "revolutionary blockchain for X" idea. Now killing it feels like admitting failure publicly.

Reality: Killing bad ideas early is a sign of competence, not failure.

3. Narrative Momentum

"The story is so good! Users will love it!"

Why it fails: A compelling narrative doesn't equal market demand. You can have a great story and zero paying customers.

Example: "AI that writes your resume while you sleep!" Sounds cool. But do people actually struggle enough with resume writing to pay $20/month? (Probably not—they do it once per year.)

4. Optimism Bias

"It'll work if I just try this one more channel / feature / pivot."

Why it fails: Every founder thinks their idea will be in the 10% that succeed. Math says 90% are wrong.

Example: Validation test fails (50 signups, threshold was 100). Instead of killing it, you think "Maybe Facebook ads are the wrong channel. Let's try Reddit." Then "Maybe Reddit users don't convert. Let's try Twitter." Endless pivots avoid the hard truth: the market doesn't want this.

5. Fear of Wasting Time

"If I kill this, the last 3 months were wasted."

Why it fails: Time is only wasted if you keep working on a bad idea. Killing it is cutting your losses, not wasting time.

Example: You've spent 100 hours. Killing now = 100 hours "lost." Continuing for 6 more months = 600 hours lost. Which is worse?

The Result: Bad Ideas Become Zombies

Founders keep zombie ideas alive long past their expiration date:

Validated weakly → "We'll iterate"
Built the MVP → "We'll find product-market fit"
Launched to crickets → "We'll improve marketing"
Year 2 → "We'll pivot the positioning"
Year 3 → Burn out, shut down

The opportunity cost is devastating. Every month on a zombie idea is a month not working on a potential winner.

The Solution: Systematic Kill/Greenlight Reviews

Our approach (Rule 007: 007-core-portfolio-kill-greenlight.mdc) makes these decisions systematic, not emotional.

The Core Principles

1. Expected Value > Sunk Cost

Formula: Expected Value = (Probability of Success) × (Impact if Successful)

Rule: Always choose the option with higher EV, regardless of past investment.

Example:

Idea A (current): 10% chance of $100K/year = $10K EV, requires 6 more months
Idea B (new): 30% chance of $200K/year = $60K EV, requires 6 months

Decision: Kill A, start B. Even though you've invested in A, B has 6x higher EV.

2. Clear Kill Criteria (No Negotiation)

Kill immediately if:

Portfolio Score < 20/40
Opportunity Score < 8.0/10
Validation thresholds missed by >30%
Founder loses conviction ("I'm forcing myself to work on this")

No exceptions. If it hits kill criteria, it's dead. No "one more week."

3. Pivot ≠ Greenlight

Pivot means: Good problem, wrong solution. Try a different approach to the same problem.

Greenlight means: Good problem, promising solution. Build it.

Kill means: Wrong problem or unsolvable. Stop entirely.

Example:

Meal planner for busy parents: Scores 22/40, validation weak
Kill? If the problem is "meal planning is annoying but not urgent," yes—kill. Wrong problem.
Pivot? If the problem is "meal planning for dietary restrictions is urgent," maybe pivot to that niche.

Most "pivots" are just delayed kills. Be honest.

4. Weekly Portfolio Reviews (Not Ad-Hoc)

Schedule: Every Friday at 4 PM (or similar ritual)

Purpose: Review all active ideas, make kill/greenlight/pivot decisions

Participants: Solo founder (you) or small team (founder + key stakeholders)

Duration: 30-60 minutes

Format: Open the Hub, review filters, make decisions

Why it matters: Scheduled reviews prevent drift. You don't work on ideas for months without checkpoints.

5. Opportunity Backlog (Not Everything Gets Started)

Concept: Not every scored idea moves to validation. Some stay in the backlog (scored but not yet started).

Why: Limited bandwidth. Better to do 2 ideas exceptionally than 5 mediocrely.

Rule: Only start new ideas if:

All higher-scoring ideas are complete or killed
You have bandwidth (time, budget, mental energy)

Backlog management:

Top Priority (score ≥ 30): Start immediately
High Priority (score 25-29): Start when bandwidth available
Medium Priority (score 20-24): Backlog (may never start)
Low Priority (score < 20): Kill immediately

The Kill/Greenlight Review: Step-by-Step

Here's exactly how we run weekly portfolio reviews.

Step 1: Open the Hub

Navigate to /hub and load all projects.

View:

All Ideas: 16 total
Top Priority (≥30): 7 ideas
Proceed: 7 ideas (passed validation)
Pivot: 9 ideas (failed validation or below threshold)

Step 2: Review Active Ideas (Currently in Discovery/Validation/Build)

For each active idea, check:

Discovery phase:

Is Opportunity Score ≥ 8.0? (If no → kill or pivot)
Are all 4 discovery docs complete? (If no → complete or kill)
Do we still believe in the problem? (If no → kill)

Validation phase:

Are we on track to hit thresholds? (If no → analyze why)
Is velocity acceptable? (If behind pace → boost effort or pivot)
Are we learning what we expected? (If no → pivot or kill)

Build phase:

Is the feature shipping on time? (If no → why? Remove scope or delay)
Are we still excited about this? (If no → kill)

Decision:

Greenlight (continue): On track, hitting metrics, team is motivated
Pivot (iterate): Missed some metrics, but problem is still valid
Kill (stop immediately): Missed metrics badly, or lost conviction

Step 3: Review Paused Ideas (In Backlog)

For each backlog idea, check:

Has anything changed? (Market shift, new competitor, new insight)
Should we re-prioritize? (Score improved due to new data)
Should we kill? (Market no longer exists, scored lower on reflection)

Decision:

Promote (move to active): Higher priority now
Keep in backlog: Still valid, just lower priority
Kill: No longer worth pursuing

Step 4: Review Killed Ideas (Recently Stopped)

For each killed idea, ask:

Did we learn anything valuable? (Document lessons)
Any salvageable assets? (Code, designs, research → repurpose)
Any ideas for other niches? (Same solution, different market)

Purpose: Extract value from failures. Killed ideas aren't wasted if you learn.

Step 5: Triage New Ideas (Scored This Week)

For each newly scored idea:

Score ≥ 30: Add to "Active" (start discovery immediately)
Score 20-29: Add to "Backlog" (start when bandwidth available)
Score < 20: Kill immediately (don't waste time)

Step 6: Update Hub Verdicts

For any decision made (kill, pivot, greenlight):

Update PORTFOLIO-SCORE-<slug>.md with new verdict
Update OPPORTUNITY-<slug>.md or RESULTS-<slug>.md with rationale
Hub will reflect updated verdict on next refresh

Step 7: Communicate Decisions (If Team)

For teams (not solo founders):

Share decisions in Slack / team meeting
Explain rationale (don't just say "we're killing X")
Celebrate kills (not just greenlights)—killing bad ideas is a win

Real-World Example: Friday Review (Dec 2, 2025)

Let me walk you through a real review session.

Context: 16 ideas in portfolio, 3 active (1 discovery, 1 validation, 1 build), 7 backlog, 6 killed

Active Idea 1: Habit Tracker (Validation Phase)

Status:

Score: 32/40 (Top Priority)
Opportunity Score: 8.5/10 (above threshold)
Validation: 104 signups (threshold: 100 ✅), 11.5% prepayment (threshold: 10% ✅)
Velocity: Hit thresholds in week 3 (ahead of 4-week plan)

Discussion:

Founder: "Validation passed. Users are paying. Feedback is enthusiastic. Should we greenlight for build?"
Review: Metrics clear, conviction high, brand system blueprint complete, dev quality plan ready

Decision: GREENLIGHT → Move to build phase

Action: Update verdict to PROCEED, start PRD/ADR next week

Active Idea 2: AI-Powered Meal Planner (Validation Phase)

Status:

Score: 22/40 (Medium Priority—barely above kill threshold)
Opportunity Score: 7.2/10 (below 8.0 threshold)
Validation: 200 signups (threshold: 500 ❌), 4% prepayment (threshold: 10% ❌)
Velocity: Week 4, significantly behind pace

Discussion:

Founder: "We're only 40% to signup threshold and prepayment conversion is half what we need. But I've already built the waitlist, wrote the copy, spent $300 on ads."
Review: Sunk cost talking. Metrics are clear—this isn't working. Opportunity Score was marginal to begin with (7.2). Users are lukewarm (4% prepayment = "maybe I'll try it" not "take my money").

Options:

Continue: Spend 4 more weeks, maybe hit 500 signups (still won't fix 4% conversion)
Pivot: Find a higher-urgency niche (e.g., meal planning for diabetics, not busy parents)
Kill: Admit this problem isn't urgent enough, move to higher-EV idea

Decision: KILL → Not worth 4 more weeks

Rationale:

Opportunity Score below threshold (7.2 < 8.0)
Validation missed by >50% (200/500 signups, 4%/10% conversion)
Founder conviction wavering ("I'm forcing myself to work on this")
Better idea in backlog (fintech SaaS, score 28/40, higher urgency)

Action: Update verdict to KILL, document lessons in RESULTS doc, archive Lindy automations, move to next idea

Active Idea 3: Project Management for Freelancers (Discovery Phase)

Status:

Score: 28/40 (High Priority)
Opportunity Score: Not yet calculated (still in discovery)
Discovery: 3/4 docs complete (missing OPPORTUNITY doc)

Discussion:

Founder: "Pain signals are strong (freelancers struggle with client communication). JTBD is clear. Working on opportunity analysis now."
Review: On track, no red flags yet

Decision: CONTINUE → Finish discovery, re-review next week when Opportunity Score is ready

Action: Complete OPPORTUNITY doc by Friday

Backlog Idea 1: Fintech SaaS for Small Businesses (Score 28/40)

Status:

Scored last month, sitting in backlog
Not started yet (bandwidth was full)

Discussion:

Founder: "Now that meal planner is killed, I have bandwidth. This scored 28/40—should I start discovery?"
Review: High score, good problem (cash flow management for SMBs), founder has domain expertise

Decision: PROMOTE → Move to active, start discovery next week

Action: Kick off NICHE-INTEL research (Manus), target 4 discovery docs in 2 weeks

Summary of Decisions

Habit Tracker: GREENLIGHT (validation passed)
Meal Planner: KILL (missed validation thresholds, below Opp Score)
Freelancer PM: CONTINUE (discovery on track)
Fintech SaaS: PROMOTE (moved from backlog to active)

Time spent: 45 minutes

Value: Killed 1 mediocre idea (saving 4+ months), greenlighted 1 winner, started 1 high-potential idea

How the Hub Enables This Ritual

The Hub makes kill/greenlight reviews visual and systematic.

Hub View (Friday Review)

All Ideas (16):

┌─────────────────────────────────────────┐
│ Filters:                                │
│ [All: 16]  [⭐ Priority: 7]             │
│ [✅ Proceed: 7]  [⚠️ Pivot: 9]          │
└─────────────────────────────────────────┘

📊 Habit Tracker (Score: 32/40 ⭐⭐⭐⭐⭐)
   Verdict: ✅ PROCEED
   Validation: 7/7 docs ✅
   
📊 Meal Planner (Score: 22/40 ⭐⭐⭐)
   Verdict: ⚠️ PIVOT (behind pace)
   Validation: 5/7 docs (results weak)
   
📊 Freelancer PM (Score: 28/40 ⭐⭐⭐⭐)
   Verdict: (pending Opp Score)
   Discovery: 3/4 docs
   
📊 Fintech SaaS (Score: 28/40 ⭐⭐⭐⭐)
   Verdict: (not started)
   Backlog

... (12 more ideas)

What the Hub shows:

Score + stars: Instantly see priority (30+ = top, 20-29 = high, < 20 = kill)
Verdict badge: Green (proceed), yellow (pivot), red (kill)
Doc completion: See progress (3/4 discovery, 7/7 validation)
Filters: Click "Pivot" to see all struggling ideas at once

Why it matters:

No spreadsheets to maintain (Hub auto-discovers from files)
No memory required (all data visible)
No debate about facts (scores/verdicts are canonical)

Making Kill Decisions Visible

When you kill an idea:

Update PORTFOLIO-SCORE-<slug>.md:

### Verdict: ❌ **KILL**

Rationale:
- Opportunity Score: 7.2/10 (below 8.0 threshold)
- Validation: 200/500 signups (40%), 4%/10% prepayment
- Founder conviction: Low (forcing myself to work on this)
- Decision: Not worth 4 more weeks. Moving to higher-EV idea.

Date: 2025-12-02

Hub reflects updated verdict:

📊 Meal Planner (Score: 22/40 ⭐⭐⭐)
   Verdict: ❌ KILL (below threshold)
   Validation: 5/7 docs (results weak)

Filter by "Pivot" shows killed ideas:

[⚠️ Pivot: 9]  ← Includes killed ideas

Benefit: Six months later, you can review why you killed it. No tribal knowledge. The decision is documented.

Resisting Sunk Cost Fallacy: The Mental Models

Here are the mental models that help us kill ideas ruthlessly.

1. "What Would I Do If I Started Today?"

Question: If I had zero investment in this idea, would I start it now given current data?

If no: Kill it. You're only continuing because of sunk cost, not because it's a good idea.

Example:

Meal planner: 3 months invested, $300 ad spend, 200 signups
Ask: If I knew right now that this idea would take 4 more weeks and deliver 500 signups at 4% conversion, would I start it?
Answer: No. I'd start the fintech SaaS (score 28, higher urgency) instead.
Decision: Kill the meal planner.

2. "Opportunity Cost is Real Cost"

Concept: Every hour on Idea A is an hour not on Idea B.

Formula: Cost of continuing A = (EV of B - EV of A) × Time

Example:

Idea A (meal planner): 10% chance × $50K/year = $5K EV, requires 6 months
Idea B (habit tracker): 40% chance × $150K/year = $60K EV, requires 6 months
Opportunity cost of continuing A = ($60K - $5K) × 6 months = $55K foregone

Decision: Kill A, start B immediately.

3. "Celebrate Kills, Not Just Wins"

Mindset shift: Killing a bad idea is a win. It's evidence of good judgment, not failure.

Ritual: When you kill an idea, celebrate it:

"Killed the meal planner today. Saved 4 months. Starting fintech SaaS next week."
Share with team / accountability partner
Document lessons learned

Why it works: Reframes kills as positive decisions, not admissions of failure. Makes future kills easier.

4. "Time Invested ≠ Future Value"

Concept: Past time is sunk. The question is always "what's the best use of my next hour?"

Example:

You've built 60% of an MVP (300 hours)
Validation is weak (users lukewarm)
Finishing would take 200 more hours

Sunk cost thinking: "I've already invested 300 hours. I have to finish."

Rational thinking: "The next 200 hours are worth more on a better idea."

Decision: Kill it. Those 300 hours are gone either way.

Practical Application: How to Implement This

Here's how to add kill/greenlight reviews to your workflow:

Step 1: Define Kill Criteria

Write down clear, non-negotiable kill criteria:

Portfolio phase:

Score < 20/40 → Kill immediately

Discovery phase:

Opportunity Score < 8.0 → Kill or pivot

Validation phase:

Miss thresholds by >30% → Kill or pivot
Founder conviction < 7/10 → Kill

Build phase:

Shipping delays > 2x estimate → Kill or descope
User feedback consistently negative → Kill

Make this visible (pin in Slack, print on wall, add to Hub).

Step 2: Schedule Weekly Reviews

Pick a time: Friday 4 PM (or similar)

Set recurring calendar event: "Portfolio Review" (30-60 min)

Invite participants: Solo founder (you) or key stakeholders

Format:

Open Hub
Review active ideas (greenlight / pivot / kill)
Review backlog (promote / keep / kill)
Triage new ideas (start / backlog / kill)
Update verdicts in Cursor
Share decisions (if team)

Step 3: Use the Hub for Visibility

Before review:

Hub shows all ideas with scores, verdicts, doc completion
Filter by "Pivot" to see struggling ideas

During review:

Click project cards to review docs
Reference Opportunity Scores, validation results
Make decisions based on visible data

After review:

Update verdict in PORTFOLIO-SCORE-<slug>.md
Hub reflects changes on next load

Step 4: Document Every Kill Decision

When you kill an idea, write why in the portfolio score doc:

### Verdict: ❌ **KILL**

Rationale:
- [Specific metrics that failed]
- [Founder conviction level]
- [What you learned]
- [What you're starting instead]

Date: [YYYY-MM-DD]

Purpose: Six months later, you'll remember why you killed it. Prevents re-litigating the same decision.

Step 5: Celebrate Kills

When you kill an idea:

Tweet about it (if public builder)
Share in founder community
Tell accountability partner
Document lessons learned

Framing: "Killed [idea] today. Validated that [problem] isn't urgent enough. Learned [lesson]. Starting [better idea] next week."

Why it matters: Reframes kills as smart decisions, not failures. Makes future kills easier.

Real-World Results

Since implementing weekly kill/greenlight reviews (6 months ago):

Portfolio health:

Started with 23 ideas (many mediocre)
Killed 6 immediately (score < 20)
Killed 3 after discovery (Opp Score < 8.0)
Killed 2 after validation (missed thresholds)
Greenlit 4 for build
Current portfolio: 16 ideas, 7 proceed, 9 pivot/killed

Time saved:

11 killed ideas × ~4 months each = ~44 months of wasted effort avoided
Focused resources on 4 high-EV ideas instead

Success rate:

4 greenlit ideas, 3 in production, 1 in build
75% success rate (vs. industry 10% without systematic portfolio management)

Confidence:

Zero regrets about killed ideas (data was clear)
Zero zombie ideas (everything active is high-conviction)

Trade-Offs and Limitations

Systematic kill/greenlight reviews aren't free:

Requires Discipline:

Easy to skip weekly reviews when busy
Easy to rationalize "one more week" instead of killing
Need commitment to follow kill criteria

Can Feel Harsh:

Killing ideas you've invested in is emotionally hard
Team members might be attached to ideas you kill
Need to communicate rationally, not emotionally

Opportunity Cost of Reviews:

30-60 minutes/week on reviews = time not coding/building
For solo founders sprinting, might feel like overhead

Doesn't Eliminate Risk:

Systematic reviews reduce bad decisions, don't eliminate them
Can still greenlight an idea that fails
Kill criteria are guidelines, not guarantees

When to Skip It

Don't implement systematic reviews if:

You're only working on 1 idea (no portfolio to manage)
You're in pure exploration mode (pre-portfolio scoring)
You're validating multiple approaches to the same problem (let data accumulate before killing)

But if you have 3+ ideas at various stages, reviews pay for themselves immediately.

Takeaways

Here's what to remember about kill/greenlight reviews:

Expected Value > Sunk Cost: Always choose the higher-EV option, regardless of past investment.
Clear Kill Criteria: Score < 20, Opp Score < 8.0, validation miss >30%, low conviction → Kill immediately.
Weekly Reviews (Not Ad-Hoc): Schedule recurring 30-60 min reviews. Review active ideas, backlog, new ideas.
Hub Enables Visibility: Scores, verdicts, doc completion all visible. No debate about facts.
Document Every Kill: Write why in portfolio score doc. Prevents re-litigation.
Celebrate Kills: Killing bad ideas is a win. Reframe as smart decisions, not failures.
Opportunity Cost is Real: Every hour on Idea A is an hour not on Idea B. Choose wisely.

What's Next?

We're continuing to refine kill/greenlight reviews:

Automated kill alerts: "Meal planner is 50% behind pace—schedule kill review"
EV calculator: Input scores/validation → Get expected value comparison
Kill lessons database: Extract patterns from killed ideas (common failure modes)

Most founders can't kill ideas because of sunk costs. We make tough decisions systematically, not emotionally.

If you're managing multiple ideas, schedule weekly reviews. Use clear kill criteria. Let the Hub show you the data. Make decisions based on expected value, not sunk cost.

What if you could kill bad ideas confidently and focus resources on winners? That's the promise of systematic portfolio management.