This Is Not a Resolution
Let's be clear about what we're doing here.
An experiment is not a commitment. Not a goal. Not a promise to yourself. Not something you can fail at.
An experiment is a question with legs. You're not trying to change. You're trying to learn.
"What happens if I pause for 3 seconds before responding?" "What happens if I name the feeling out loud?" "What happens if I leave my phone in another room?"
That's it. You run the experiment. You see what happens. You get data.
The pattern might not change at all. That's fine—you learned something. The experiment might feel stupid and pointless. Also fine—data. You might forget to run it entirely. Still data.
You cannot fail an experiment. You can only get information.
The Anatomy of a Good Experiment
A good 5% experiment has five parts:
1. The Weak Point
Where in your loop are you intervening? (If you haven't identified this, go back to "Where to Interrupt.")
Be specific. Not "somewhere in my anxiety pattern" but "the gap between noticing my chest tighten and the thought 'something's wrong.'"
2. The Trigger
What cues you to run the experiment?
This should be something noticeable. A physical sensation. An external situation. A specific thought. Something that tells you: "This is the moment."
Bad trigger: "When I'm anxious" (too vague, you might not notice until it's too late)
Good trigger: "When I notice my shoulders at my ears" (specific, physical, observable)
3. The Action
What's the smallest possible thing you'll do?
Remember: embarrassingly small. If it requires sustained effort, willpower, or major disruption, it's too big. Scale down until it almost feels pointless.
Too big: "I'll do a 10-minute breathing exercise"
Right size: "I'll take one breath"
Too big: "I'll journal about my feelings"
Right size: "I'll say one word that names what I'm feeling"
Too big: "I'll set a boundary"
Right size: "I'll pause for 5 seconds before responding"
4. The Duration
How long are you running this experiment?
Not forever. Not "until it works." A defined period.
Recommended: 7 days for most experiments. Short enough to stay focused, long enough to get some data.
For very small experiments, even 3 days can work. For bigger pattern shifts, 2 weeks. But always bounded.
5. What You're Tracking
What counts as data?
You're not tracking "did it work" (too binary). You're tracking what happened.
- Did I remember to try the experiment?
- What did I notice in the moment?
- What happened next in the loop?
- How did it feel?
- What surprised me?
The Experiment Formula
When I notice [specific trigger], I will [tiny action], for [duration], and I'll track [what I noticed].
Example: "When I notice the urge to check my phone during work, I will say 'there's the urge' out loud, for 7 days, and I'll track whether saying it changed anything."
Example Experiments
Here are some properly-sized experiments for common patterns:
Procrastination Loop
Weak point: The moment between feeling overwhelmed and reaching for distraction
Experiment: When I notice overwhelm about a task, I will write down one sentence describing what feels hard about it. 7 days. Tracking: Did I do it? Did anything shift?
People-Pleasing Loop
Weak point: The gap between hearing a request and responding
Experiment: When someone asks me to do something, I will wait 5 seconds before answering. Just silence. 7 days. Tracking: What did I notice in the silence? Did I still say yes?
Anxiety Spiral
Weak point: The body sensation before the catastrophic thought
Experiment: When I notice chest tightening, I will put my hand on my chest and say "that's anxiety" before doing anything else. 7 days. Tracking: Did I catch it? What happened to the thought?
Burnout Loop
Weak point: The "I should push through" thought
Experiment: When I notice "I should push through," I will check: am I above or below 50% energy? Just notice, don't change anything. 7 days. Tracking: What did I learn about when I push?
Comparison Spiral
Weak point: The moment of opening social media
Experiment: Before opening Instagram, I will say out loud why I'm opening it. 5 days. Tracking: What reasons did I give? Did naming it change whether I opened it?
How to Track (Without Making It a Chore)
Tracking is important. But if tracking becomes another overwhelming system, you won't do it.
Option 1: The One-Word Check-In
At the end of each day, write one word about the experiment. That's it.
"Forgot." "Tried." "Weird." "Nothing." "Interesting."
One word is enough to jog your memory later.
Option 2: The Voice Note
When the experiment happens (or doesn't), record a 15-second voice note to yourself. Stream of consciousness. Don't edit.
"Just noticed the urge to scroll, said 'there's the urge,' still scrolled but it felt different, I don't know why."
Option 3: The Tally
If your experiment is very simple, just keep a tally. Did I remember? Yes/No. That's data.
Day 1: Yes | Day 2: No | Day 3: Yes | Day 4: Yes | Day 5: No
Option 4: The Pattern Log
If you're using Unloop, log when you ran the experiment. Brief notes optional. The tool tracks for you.
What's your tracking style?
Pick the one that feels easiest, not the one that feels most thorough. The best tracking system is the one you'll actually use.
What Counts as Success?
Here's the mindset shift:
Old frame: Success = the pattern changed
New frame: Success = I ran the experiment and noticed what happened
If you ran the experiment even once and paid attention, it worked. You have data you didn't have before.
Some of that data might be:
- "The intervention didn't do anything noticeable"
- "I couldn't catch the moment in time"
- "The experiment was too hard to remember"
- "I felt resistance to even trying"
All of that is useful. It tells you something about your pattern, your weak point, or your experiment design.
"Nothing happened" is a finding, not a failure.
Common Problems (and What They Mean)
"I keep forgetting to do it"
Your trigger isn't salient enough, or there's no external reminder.
Try: Add a physical cue. A rubber band on your wrist. A sticky note. A phone alarm. Something that interrupts autopilot.
"I remember right after the moment passes"
The pattern is faster than your awareness. Normal.
Try: This is still progress. "Remembering after" becomes "remembering during" which becomes "remembering before." Keep going. You're building the muscle.
"I did it but nothing changed"
The experiment might be too small to see results yet. Or you might need more repetitions. Or this isn't the right weak point.
Try: Run it for the full duration before concluding. If still nothing, try a different weak point or a slightly different experiment.
"It felt stupid and pointless"
Good experiments often feel this way. The smallness is the point.
Try: Do it anyway. "Feeling stupid" isn't data about whether it works. It's just discomfort with doing something different.
"I feel resistant to even trying"
Interesting. What's the resistance protecting?
Try: Get curious about the resistance instead of pushing through it. The resistance might be more useful data than the experiment itself.
"The pattern got worse when I paid attention to it"
Attention can temporarily amplify patterns. This is normal and usually passes.
Try: Keep going gently. You're not doing damage—you're just seeing what was already there. The heightened awareness often settles.
What Experiments Are
Questions with legs. Data collection. Curiosity in action. Ways to learn about your pattern without having to fix it.
What Experiments Aren't
Commitments. Goals. Tests you can fail. Proof that you can or can't change. Judgments about your willpower.
After the Experiment: What Now?
Your 7 days (or whatever duration) is up. Now what?
Step 1: Review the Data
Look at whatever you tracked. Not to judge—to learn.
- How often did you remember?
- What did you notice when you ran it?
- Did anything shift, even slightly?
- What surprised you?
Step 2: Ask the Key Questions
Was this the right weak point? If you consistently couldn't catch the moment, the weak point might be too fast or too automatic. Try another spot in the loop.
Was the experiment the right size? If you rarely did it, maybe it was too big. If it felt meaningless, maybe it was appropriately small (that's fine) or maybe a slightly bigger version would give more data.
What did I learn about the pattern? Even "unsuccessful" experiments teach you something. What do you know now that you didn't know before?
Step 3: Decide What's Next
You have options:
Run it again. 7 more days of the same experiment. More data, more reps, more chance to catch the moment.
Adjust and run. Same weak point, slightly different experiment. Tweak based on what you learned.
Try a different weak point. This one wasn't accessible enough. Pick another spot in the loop.
Expand by 5%. The experiment worked. Make it slightly bigger. Not a lot—just 5% more.
Rest. Take a break from experimenting. Let what you learned settle. Not every week needs an active experiment.
The Long Game
One experiment isn't going to transform your pattern. Neither is five.
This is a practice. Like meditation, or exercise, or any other thing that works slowly over time.
What you're building:
- Awareness of when the pattern runs
- Ability to catch it earlier
- Tolerance for doing something different
- Evidence that the loop isn't as fixed as it feels
- Gradually widening cracks in the automatic sequence
Patterns that took years to form don't dissolve in a week. But they can loosen. One tiny experiment at a time.
Ready to design your experiment? Map your pattern, find the weak point, and run your first 5% test.
Start MappingQuick-Start Template
Don't overthink it. Just fill in:
My pattern: ________________
The weak point I'll target: ________________
My trigger (how I'll know it's the moment): ________________
My tiny action: ________________
Duration: ________________
How I'll track: ________________
What I'm curious about: ________________
That's enough. Go run it.
Experiments aren't about fixing yourself. They're about getting curious about what's already happening—and seeing what's possible.
Try ItRemember
An experiment is not a commitment. It's a question with legs. You're not trying to change—you're trying to learn. Keep it embarrassingly small. Track without turning it into a chore. "Nothing happened" is still data. You cannot fail an experiment, you can only get information. Run it, notice what happens, adjust, repeat. Patterns loosen slowly, at weak points, through accumulated tiny interventions. One experiment at a time.