Running Energy Experiments

You've read that morning exercise boosts energy. You've heard that cutting caffeine after noon improves sleep. You've seen advice about meal timing, hydration, and stress management.

But does any of it work for you?

The only way to know is to run experiments. Not vague "I'll try this for a bit" experiments—real, structured tests that give you actual answers.

Here's how to run energy experiments that actually tell you something.

Why You Need Personal Experiments

Generic advice has a problem: it's generic. "Exercise improves energy" might be true on average, but averages don't help you if you're an exception.

Consider caffeine. General advice says cut it by early afternoon. But some people can have espresso at 6pm and sleep fine. Others need to stop by 10am. You won't know which you are until you test it.

Key Insight: Your body is the laboratory. Generic studies tell you what works for most people. Experiments tell you what works for you. Track what you control, then test what matters.

The Single-Variable Rule

The most important principle of personal experimentation: change one thing at a time.

When you change multiple variables simultaneously:

• You can't know which change helped (or hurt)
• Effects might cancel each other out
• You'll either credit the wrong change or miss what actually worked

Bad experiment: "I'll go to bed earlier, cut caffeine, and start exercising." Good experiment: "I'll move my bedtime 30 minutes earlier for two weeks. Everything else stays the same."

This feels slow. It is slow. It's also the only way to get real answers.

Anatomy of a Good Energy Experiment

1. Hypothesis

Start with a clear, testable prediction:

Vague: "Better sleep will help my energy." Specific: "If I get in bed by 10:30pm instead of 11:30pm, my morning energy rating will improve by at least 1 point."

Good hypotheses are:

• Specific (exact change, exact expected outcome)
• Measurable (you can track both input and outcome)
• Based on observation (your data suggests this connection)

2. Baseline

You need comparison data. What's your current state before the change?

From your baseline data, identify:

• Current average for the input (e.g., average bedtime: 11:30pm)
• Current average for the outcome (e.g., morning energy: 5.2/10)
• Normal variability (so you know what counts as real change)

Without baseline, you're comparing to nothing.

3. Protocol

Define exactly what you're changing:

Element	Example
What changes	Bedtime
Specific change	From ~11:30pm to 10:30pm
How long	Two weeks minimum
What stays the same	Everything else (caffeine, movement, meals)
What you track	Bedtime (actual), morning energy rating

Write this down. Refer to it when you're tempted to make other changes.

4. Duration

How long should an experiment run?

Minimum: Two weeks

• Gives you enough data points (14)
• Accounts for daily variation
• Includes weekdays and weekends

Better: Three weeks

• More robust data
• Pattern has time to establish
• Better accounts for weekly rhythm

For bigger changes: Four weeks or longer

• Sleep schedule shifts need time
• Body adaptation takes time
• More certainty in results

5. Analysis

At experiment end, compare:

• Your input average during experiment vs. baseline
• Your outcome average during experiment vs. baseline
• Whether the change meets your hypothesis

Did you actually make the change? Did the outcome actually shift? By how much?

Running Your First Experiment

Let's walk through a complete example.

The Observation

Looking at your energy trends, you notice that your highest-energy days all followed nights when you got in bed before 10:30pm. Days following later bedtimes had lower energy.

The Hypothesis

"If I consistently get in bed by 10:30pm, my average morning energy will increase from 5.2 to at least 6.0."

The Baseline

From the past three weeks:

• Average bedtime: 11:32pm
• Average morning energy: 5.2/10
• Range: 4-7 energy, 10pm-12:30am bedtime

The Protocol

Change: Get in bed by 10:30pm every night Duration: Two weeks (14 nights) Track: Actual bedtime, morning energy (1-10) Keep constant: Caffeine cutoff (2pm), movement pattern, meal timing

The Execution

Week 1:

• Day 1: Bed at 10:25pm. Morning energy: 6
• Day 2: Bed at 10:40pm. Morning energy: 5
• Day 3: Bed at 10:20pm. Morning energy: 7
• ...

Week 2:

• Continue logging
• Note any disruptions or unusual circumstances

The Analysis

After two weeks:

• Average bedtime during experiment: 10:28pm (vs. 11:32pm baseline)
• Average morning energy: 6.1 (vs. 5.2 baseline)

Result: Hypothesis confirmed. Earlier bedtime increased morning energy by nearly 1 point.

The Decision

This experiment shows a clear benefit. You can:

• Maintain the change (keep 10:30pm bedtime)
• Run a follow-up experiment (does 10pm work even better?)
• Move to testing another variable

Types of Energy Experiments

Timing Experiments

Testing when you do something:

Experiment	Change	Outcome to Track
Bedtime shift	Move 30-60 min earlier/later	Morning energy, sleep quality
Caffeine cutoff	No caffeine after X time	Sleep onset, next-day energy
Meal timing	Eat dinner earlier/later	Evening energy, sleep quality
Exercise timing	Morning vs. evening workout	Daily energy pattern

Addition Experiments

Testing if adding something helps:

Experiment	Addition	Outcome to Track
Morning movement	10-min walk before work	Morning and afternoon energy
Afternoon break	15-min midday walk	Afternoon slump severity
Hydration	Track and increase water	Energy stability
Morning sunlight	10 min outside after waking	Alertness, evening sleep

Removal Experiments

Testing if removing something helps:

Experiment	Removal	Outcome to Track
Late caffeine	No caffeine after noon	Sleep quality, morning energy
Screen before bed	No screens 1 hour before bed	Sleep onset, sleep quality
Late eating	Finish eating by 7pm	Evening energy, sleep quality
Alcohol	No alcohol for 2 weeks	Sleep quality, morning energy

Consistency Experiments

Testing if doing something regularly helps:

Experiment	Consistency Target	Outcome to Track
Sleep schedule	Same bedtime +/- 30 min	Overall energy stability
Movement frequency	Daily vs. sporadic	Baseline energy level
Meal rhythm	Consistent meal times	Energy crashes

How to Know If It's Working

Statistical vs. Meaningful

A change from 5.2 to 5.4 average energy might be statistically "real" but practically meaningless. A change from 5.2 to 6.2 is meaningful.

Ask:

• Is the change noticeable in daily life?
• Is it greater than your normal variation?
• Would you continue the change based on this result?

Signal vs. Noise

Your energy varies naturally. How do you know a change is real?

Signs it's signal:

• Consistent improvement across most days
• Effect size larger than your normal day-to-day variation
• Pattern holds through weekdays and weekends

Signs it's noise:

• Some better days, some worse
• Improvement within your normal range
• Confounding factors during the experiment

When Experiments Fail

Failed experiments are valuable:

• You learned what doesn't work for you
• You can eliminate that variable from consideration
• You can move on to testing something else

Common reasons experiments fail:

• You didn't actually make the change consistently
• The variable doesn't affect your energy
• Other confounding factors interfered
• The duration was too short

Advanced Experiment Strategies

The Reversal Test

After a successful experiment, try reversing the change:

1. Change input for two weeks (experiment)
2. Return to original behavior for two weeks (reversal)
3. Implement the change again for two weeks (confirmation)

If energy drops during reversal and rises again during confirmation, you have strong evidence.

The Graduated Test

For bigger changes, test incrementally:

• Week 1-2: Bedtime at 11:00pm (30 min earlier)
• Week 3-4: Bedtime at 10:30pm (60 min earlier)
• Week 5-6: Bedtime at 10:00pm (90 min earlier)

This shows where diminishing returns begin.

The Combination Test

After establishing individual effects, test combinations:

1. First, confirm caffeine cutoff at noon helps (single variable)
2. Then, confirm earlier bedtime helps (single variable)
3. Finally, test both together

Only combine variables you've already tested individually.

Experiment Documentation

Keep records of your experiments. For each one, track:

Field	What to Record
Start date	When the experiment began
Hypothesis	What you expected to happen
Protocol	Exactly what changed
Baseline	Your pre-experiment averages
Results	Your during-experiment averages
Outcome	Confirmed, refuted, or inconclusive
Notes	Any confounding factors or observations

This becomes your personal playbook. When energy dips in the future, you can refer back to what worked.

Common Mistakes in Energy Experiments

Changing Multiple Variables

We said it before, but it bears repeating. One change at a time.

Not Tracking the Actual Input

If you're testing bedtime, track your actual bedtime—not your target. You need to know what you actually did, not what you intended.

Abandoning Too Early

Give experiments the full duration. Don't quit after three days because "nothing is happening." Patterns take time.

Ignoring Confounding Factors

If you get sick during your experiment, note it. If work is unusually stressful, note it. Context matters.

Expecting Perfection

Real experiments in real life aren't laboratory-clean. You'll miss some days. That's okay. Look at the overall pattern.

The Trendwell Approach

Trendwell supports experiments by:

• Making it easy to track the inputs you're testing
• Showing your outcomes over time
• Revealing correlations between inputs and outcomes
• Helping you see what's actually changing

You don't need complicated software. You need consistent data and a clear protocol.

Building Your Experiment Queue

Based on your baseline and trends, create a prioritized list of experiments to run:

1. Most impactful first: Start with the input most likely to matter
2. Easiest first: Start with changes you can actually sustain
3. One at a time: Complete each before starting the next

A typical queue might be:

1. Caffeine cutoff experiment (2 weeks)
2. Bedtime experiment (2 weeks)
3. Morning movement experiment (2 weeks)

That's six weeks of learning about your energy. And at the end, you'll know what actually works for you—not what works for some study population.

Next Steps

• Review your baseline data: What correlations do you see?
• Form one hypothesis: Based on your data, what seems likely to help?
• Define a protocol: What exactly will you change? For how long?
• Start your first experiment: Begin tracking today
• Read: Finding Your Energy Correlations
• Read: Track Energy Inputs, Not Fatigue

Your energy is personal. The only way to understand it is to run personal experiments. Start with one change, track it properly, and let your data tell you what works.

---

Last updated: January 2026