Head of R&D for My Own Life

Running a few things at once now. Kill some. Double down on others. Test constantly.

Feels less like founder. More like head of R&D for my own life.

Most advice doesn't fit anymore. The playbooks assume you're all in on one thing.

Nobody writes about this version.

The All-In Myth

Every startup playbook says the same thing: Focus. Pick one idea. Go all in. Everything else is distraction.

Paul Graham: "The biggest mistake founders make is not focusing enough."

YC: "Do one thing exceptionally well."

Naval: "Specific knowledge is found by pursuing your genuine curiosity."

They're not wrong. For venture-backed companies burning $300K/month, focus is survival.

But what if you're not burning cash? What if you're profitable? What if AI lets you run multiple experiments simultaneously?

The playbooks break down.

My Current Portfolio

Here's what I'm running right now:

Active Projects:

• Portfolio website → Testing different positioning
• Newsletter acquisition → Evaluating 3 potential buys
• Content creation → Podcast + blog + Twitter
• Investment research → Building systematic processes
• ADGM setup → International tax optimization

Recent Kills:

• SimpleDirect Chat (customers could use ChatGPT)
• Enterprise sales attempts (wrong fit for bootstrap model)
• Complex trading strategies (returns worse than VOO)

Doubling Down:

• Geographic arbitrage (moving costs to India)
• AI-first development (10x productivity gains)
• Content distribution (growing fastest channel)

Why This Works Now

1. AI Changed the Economics

Used to need a team for each project. Now I need tools.

• Cursor writes code faster than I can review
• Claude handles research and analysis
• Make automates workflows
• India team scales execution

Cost per experiment: $500-2K/month vs $50K+/month

2. Portfolio Effects

Each project teaches me something useful for the others.

Newsletter research → Better content topics Content creation → Better deal sourcing
Investment analysis → Better business evaluation Tax optimization → Better profit retention

The knowledge compounds.

3. Optionality Beats Optimization

In uncertainty, having more shots on goal matters more than perfect execution on one.

Market changes fast. AI changes faster. Customer needs change fastest.

Better to test 5 things at 70% effort than perfect 1 thing that becomes obsolete.

The New Playbook

Week 1-4: Rapid Testing

• Launch 3-5 small experiments
• Spend $500-2K max per experiment
• Set clear kill criteria upfront
• Measure leading indicators, not vanity metrics

Week 5-8: Data Gathering

• Which experiments show traction?
• Which feel energizing vs draining?
• Which have natural synergies?
• Which could scale without me?

Week 9-12: Portfolio Pruning

• Kill bottom 50% (be ruthless)
• Double resources on top 25%
• Let middle 25% run on autopilot
• Start next batch of experiments

Every Quarter: Strategic Review

• What's working? Why?
• What's not? Why not?
• What new opportunities emerged?
• What should I test next?

The Sunday Night Test

My framework for keeping experiments vs killing them:

Sunday night, 9 PM. Tomorrow's Monday. How do you feel about working on this project?

Score 1-4:

• 1: Dread (kill immediately)
• 2: Neutral (probably kill)
• 3: Positive (keep running)
• 4: Excited (double down)

The $3M Partnership I Walked Away From

Contract looked perfect on paper. Recurring revenue. Brand-name client. VCs would've loved it.

Sunday Night Test: 1 out of 4.

Would've dropped my autonomy to 2. Locked me into client-dependent deliverables. No scaling potential.

I said no. Used that time to build SimpleDirect Desk instead.

The test doesn't lie.

What Nobody Tells You

1. You Need Systems, Not Passion

Passion burns out. Systems compound.

I don't love every project every day. But I love the portfolio effect. I love the optionality. I love learning what works.

2. Most Experiments Should Fail

If 80% of your experiments succeed, you're not testing risky enough ideas.

I target 30% success rate. High enough to make progress. Low enough to find breakthrough opportunities.

3. Kill Fast, Kill Often

The hardest part isn't starting experiments. It's stopping them.

Sunk cost fallacy is real. "I already spent $5K on this."

That's exactly why you should kill it. Don't spend the next $50K.

4. Energy > Revenue (Initially)

Which projects give you energy? Which drain you?

Revenue follows energy, not the other way around.

I've killed profitable projects that felt like death marches. I've kept barely-profitable projects that taught me new skills.

The Advice That Doesn't Apply

"Focus on one thing" → Works for VC-backed companies with infinite money and time pressure. Doesn't work for profitable solopreneurs with multiple interests.

"Find your niche" → Assumes markets are static. In AI-first world, niches get automated overnight. Better to be adaptable than specialized.

"Build something people want" → Right, but which people? Test multiple audiences simultaneously. Let market demand tell you where to focus.

"Overnight success takes 10 years" → Used to be true. AI compressed timelines. Now overnight success takes 10 experiments.

Who This Works For

Good fit:

• Profitable solo operators
• AI-first builders
• Geographic arbitrage practitioners
• People who get energized by variety
• Those with systems thinking

Bad fit:

• Venture-backed companies (burn rate demands focus)
• People who need deep expertise (doctors, lawyers)
• Those paralyzed by too many options
• Teams bigger than 5 people

My Next Experiments

Testing in Q1 2026:

1. Newsletter roll-up model → Buy 3-5 profitable newsletters, systematize operations
2. AI consulting for family offices → High-value, low-volume service business
3. Geographic arbitrage course → Productize my ADGM/nomad setup knowledge
4. Investment newsletter → Document my portfolio decisions publicly

Kill criteria for each:

• Revenue potential <$10K/month
• Time investment >20 hours/week
• Negative Sunday Night Test for 4 weeks
• Requires team >3 people

The Meta Game

Here's what I've learned running this portfolio approach for 18 months:

The real product isn't any individual project. It's the system that generates and evaluates projects.

Most founders optimize for one outcome. I optimize for multiple outcomes and the ability to recognize which ones matter.

The real moat isn't the current projects. It's the speed at which I can test new ones.

While competitors perfect their single approach, I'm testing the next 5 approaches.

The real defensibility isn't product features. It's operational capability.

AI tools, offshore team, tax optimization, content distribution, deal sourcing—the stack compounds.

Action Steps

If this resonates, here's how to start:

This Week:

• List all your current projects/interests
• Run Sunday Night Test on each (1-4 score)
• Kill anything scoring 1-2
• Set kill criteria for everything scoring 3-4

This Month:

• Start 2-3 small experiments ($500-2K budget each)
• Build systems for tracking and evaluation
• Find your AI tools and offshore team
• Document what you learn

This Quarter:

• Review what worked/didn't work
• Double down on top performers
• Kill bottom performers
• Plan next batch of experiments

Why I'm Writing This

Because the internet is full of "focus" advice and zero "portfolio" advice.

Because AI changed the game and nobody updated the playbooks.

Because being head of R&D for your own life is more fun than being an all-in founder.

And because maybe you're running multiple things too, wondering if you're doing it wrong.

You're not. You're just playing a different game with different rules.

The question isn't whether to focus or diversify.

The question is: What's the right portfolio size for your situation?

For me, it's 3-5 active experiments. Enough for optionality. Not so many I can't execute well.

What's yours?