How I Researched 25 Habit-Building Techniques with NotebookLM

Hey,

Last newsletter, I told you about building an app with my AI team. Today, I want to show you how I researched the psychology behind it.

Because here’s the thing. I didn’t want to just build another habit app based on one book. I wanted to find what actually works, backed by science.

So I went deep. And I found 25 different methodologies for building habits.

Let me show you how you can do the same for any research project or analysis.

The Tool That Changed My Research Game

The secret weapon: NotebookLM (notebooklm.google.com)

Think of it as your personal research assistant that only uses sources you trust. No random internet opinions. No hallucinated facts. Just your curated sources, analyzed and connected.

Primož, why is this better than ChatGPT?

When you ask a regular AI about habit formation, it pulls from everywhere. You get a mix of “psychology”, outdated studies, new studies, and random blog posts. With NotebookLM, you control the knowledge base.

3-Step Research Process

Step 1: Feed it quality sources

I started by adding scientific articles and studies about habit formation. My prompt to find sources:

“Scientific articles and studies demonstrating the most effective methods for forming sustainable habits”

You can add PDFs, Google Docs, YouTube videos, websites. The key is choosing sources you actually trust.

I used Deep Research, which finds web sources automatically. But because I wrote “scientific articles and studies about…” in my prompt, NotebookLM only added sources from scientific articles and studies.

Once it finished, I added the found sources:

Step 2: I asked for rankings

Once my sources were loaded, I asked:

“Rank all methodologies and techniques from those that demonstrate the highest probability of sustaining the new habit to those with the lowest effectiveness.”

So instead of reading 20 papers myself, AI synthesized them, found patterns across all my sources, and included references in the response.

Step 3: I went deeper on what matters

After the ranking, I dug deeper. Asked follow-up questions. Requested specific evidence. Built a complete picture.

What I Found: 25 Techniques, Ranked

Here’s the full list I discovered, from evidence-based psychology to ancient wisdom:

The Heavy Hitters (Strong Research Support):

Implementation Intentions (“If-Then” plans)
WOOP/MCII (Wish-Outcome-Obstacle-Plan)
Fogg Behavior Model + Tiny Habits
COM-B + Behavior Change Wheel
Habit Loop (Cue-Routine-Reward)

The Psychology Frameworks:

Motivational Interviewing
CBT Behavioral Activation
Self-Determination Theory
Transtheoretical Model (Stages of Change)
Identity-Based Habits
ACT (Acceptance & Commitment Therapy)

The Practical Systems:

GROW Coaching Model
Nudge Theory & Choice Architecture
Kaizen (1% improvements)
Digital Behavior Change Interventions

The Ancient Wisdom:

Mindfulness (Vipassana)
Zen/Zazen - Pranayama
Qigong/Tai Chi
Ikigai
Wu Wei (Daoism)
Yoga philosophy (Yamas & Niyamas)

The surprise?

The most effective techniques weren’t complicated. Implementation Intentions (“If it’s 7am, then I meditate for 2 minutes”) consistently outperformed complex systems.

If you click the “infographic” button:

You get a nice, yes you guessed it, infographic from all your sources 🙂:

Try This

Pick something you want to research “deeply”.

Go to notebooklm.google.com
Add trusted sources (articles, videos, documents…)
Ask: “What are the key insights across all these sources?”
Then: “Rank the approaches by effectiveness”
Finally, create an infographic

And just like that, we have another AI tool in our toolkit… To become better! 😉

What I’m Building With This Research

All 25 techniques are now influencing the design of my app that I’m building. Not just copying insights from the “Atomic Habits” book, but combining the best of behavioral science.

The Goal:

Combine the best of behavioral science into one simple tool that actually helps me change.