Experiments · Reflection · HourLife Premium
Try one change on purpose, rate it each evening, and say what you learned.
Test one change for 7 to 30 days, rate it each evening, and record your own verdict when the run ends.
-
One change, a fixed runName the change and what you expect, pick 7, 14, 21 or 30 days, and it starts. The app writes it as an “If I…, then…” hypothesis and splits the run into weeks.
-
Rated each eveningEach measure becomes a question on Today’s Evening card, answered with one tap from 1 to 5. The experiment keeps the average, week by week, with no interpretation added.
-
A verdict, in your wordsWhen you conclude, you review the numbers, say whether the hypothesis held, and choose to keep it, run a follow-on, keep observing, or drop it. Keeping it can start a habit.
Your first experiment
-
Name the change
Open Experiments and tap "+". Type the change under "WHAT ARE YOU TESTING?" and what you expect under "EXPECTED RESULT". "Start" waits for both while "Require hypothesis" is on in Settings, which it is to begin with.
-
Pick a length and a measure
Choose 7d, 14d, 21d or 30d under "RUN LENGTH", leave "Start today" on, and add a "Quality" measure such as "Sleep quality" under "MEASURE". Then tap "Start".
-
Answer it each evening
Evening mode on Today now carries an "EXPERIMENTS" card asking "How was your sleep quality today?". Tap 1 to 5 once a day, and write the week's check-in from the experiment's "Notes" tab.
Every step, in detail
- 1 Open Experiments and tap "+" to start a "New Experiment".
- 2 Under "WHAT ARE YOU TESTING?", name the change, and under "EXPECTED RESULT" say what you expect. The "PREVIEW" turns them into an "If I..., then... within N days." hypothesis.
- 3 Pick a "RUN LENGTH" of 7, 14, 21, or 30 days, and leave "Start today" on or choose a start date.
- 4 Under "MEASURE", pick "Quality", "Quantity", or "Yes/No", name what you will track, and tap "Start".
A run, day by day
An experiment starts the moment you tap Start, and its run is cut into weeks from that minute. Each evening the measures ask for a rating; each week asks for a check-in; and the list counts down to the end date. Drag through two weeks to see what each screen says along the way.
Sample · “Charge my phone outside the bedroom”, 14 days from Monday, Oct 12 at 9 pm, two Quality measures · drag the day, tap tonight’s ratings
The run
Two weeks, rated each evening
Each bar is one evening’s tap, 1 to 5. A gap is an evening with no rating; the average simply has one fewer reading.
Week 1 Oct 12 – Oct 18
- Mon12
- Tue13
- Wed14
- Thu15
- Fri16
- Sat17
- Sun18
Week 2 Oct 19 – Oct 25
- Mon19
- Tue20
- Wed21
- Thu22
- Fri23
- Sat24
- Sun25
Ends Monday, Oct 26 at 9 pm, fourteen days after the start.
Charge my phone outside the bedroom
How was your sleep quality today?
How was your morning energy today?
Charge my phone outside the bedroom
If I charge my phone outside the bedroom, then I fall asleep faster within 14 days.
Week 1Oct 12 – Oct 18
CompleteWeek 2Oct 19 – Oct 25
Active · day 2
Week 1Asleep before 11 on most nights.
Week 2Not yet written · check-in due Oct 25
Nothing ends the run for you. Past its end date the card still reads ACTIVE with “0d left”, and the Evening card keeps asking, until you choose Conclude. The list’s “Nd left” counts whole days to the end, while the Today card that appears in the last days counts days of the run, so the two can differ by one. “Write” opens the week’s check-in: how it is going, a note, and whether to continue, modify or stop. Choosing Stop there is kept with the check-in and does not end the run.
The numbers it keeps, and what it leaves to you
The Data tab averages what you logged and stops there. It draws no line through the dots and passes no verdict. An experiment you start has no baseline period, so the comparison is yours to make: one week against the next, or the run against how you remember the weeks before it.
Sample · the same run on its last day, Monday, Oct 26
Metrics · 2 tracked
Sleep qualityLatest: 5 on Oct 25 · 12 entries
5Morning energyLatest: 4 on Oct 25 · 12 entries
4“Log data” adds or corrects a value for any date.
Baseline vs now
| Metric | Baseline | Now | Δ |
|---|---|---|---|
| Sleep quality | — | 3.92 | — |
| Morning energy | — | 3.42 | — |
“No interpretation — numbers only. Draw your own conclusions.”
Timeline
Week 2Oct 19 – Oct 25
Top: Sleep quality 4.33Week 1Oct 12 – Oct 18
Top: Sleep quality 3.50
With no baseline there is no percentage, so each week shows one measure’s plain average.
Concluding, in three steps
“Conclude” in the experiment’s ⋯ menu walks through the numbers, your verdict and what happens next. Concluding files the experiment under Archived with its verdict, and can open a decision or a habit straight after.
Sample · the same run, concluded on Oct 26 · choose a verdict and what happens next
1Review the numbers
Raw data only — no interpretation yet.
Sleep quality
Morning energy
Check-ins
Wk 1
On trackWk 2
On track
2Was the hypothesis correct?
If I charge my phone outside the bedroom, then I fall asleep faster within 14 days.
3What happens next?
“Conclude experiment” writes
- ArchivedCharge my phone outside the bedroom · Concluded · Oct 26, 2026
- VerdictSupported · Kept permanently
“Graduate to a habit” is offered only with Keep permanently. With both boxes ticked, the decision opens first and the new habit follows when you close it. “Run a follow-on experiment” records the choice; it does not set up the next experiment for you.
Insights keeps the score
“Insights” in the ⋯ menu counts every experiment you have run, archived ones included, by where it stands, and tallies how your hypotheses held up.
Sample · five earlier experiments and this run, on Tuesday, Oct 20
You’ve concluded 4 experiments.
1 running, 0 waiting for review.
Pipeline
Where your experiments sit.
2 you decided to keep for good.
What you’ve found
Your hypotheses, put to the test.
- ✓Supported2
- ◐Partially supported1
- ✕Not supported1
Archived
Where the five earlier runs live.
Stopped counts the one deleted from its own page, which files it here as Abandoned.
No phone for the first hourConcluded · Sep 27, 2026
Read on paper before bedConcluded · Aug 30, 2026
Standing desk afternoonsConcluded · Jul 19, 2026
Walk after lunchConcluded · Jun 14, 2026
Cold shower finishAbandoned · May 9, 2026
Tips
- Keep it to one change and one or two measures. Two questions an evening is easy to keep up for a month.
- Rate at the same point each evening, from Evening mode, so the numbers compare like with like.
- Write the week's check-in before the week ends. Only the current week takes one; a week that passes without one reads "Missed".
- Turn off "Require hypothesis" in "Settings" to start an experiment with no expected result. It is written as "Try ... for N days." instead.
- Set a "Default duration" in "Settings" if you usually run experiments for the same length.
Questions
About Experiments day to day.
-
Do I have to measure anything?
No. "MEASURE" is optional, and an experiment with none still has its run, weekly check-ins and notes. A measure is what gives the Data tab and the conclusion numbers to show.
-
How do I log a measure each day?
Evening mode on Today shows an "EXPERIMENTS" card with one row per measure and the buttons 1 to 5, from "Poor" to "Great". One tap logs today, and tapping again changes it. To log another date or an exact number, open the experiment, go to "Data" and tap "Log data".
-
Does the app tell me whether it worked?
No. The Data tab says so: "No interpretation — numbers only. Draw your own conclusions." It shows averages, and when you conclude you choose the verdict yourself: supported, partially supported, inconclusive, or not supported.
-
Does an experiment stop by itself when the run ends?
No. Past its end date it stays on the list as ACTIVE with "0d left", and the Evening card keeps asking, until you choose "Conclude" from its ⋯ menu.
-
Where do finished experiments go?
Concluding moves an experiment to "Archived" in the ⋯ menu, with its verdict and decision. From there you can unarchive it or delete it permanently. "Delete" on an experiment's own page also moves it there, marked Abandoned.
-
Is Experiments free, and does it sync?
Experiments is part of HourLife Premium. Experiments sync through iCloud with the rest of your data, so they are on every device signed in to your account.
Works with
Where Experiments hands off to the rest of HourLife.
-
Today
Evening mode shows an "EXPERIMENTS" card that asks each measure's question with a 1 to 5 rating, and Today can carry an active experiments widget.
-
Habits
You can link an experiment to a habit in "Edit", and a kept experiment can graduate into a new habit named after it.
-
Decisions
When you conclude, "Create a decision record" opens a new decision asking "Should I keep ...?" with the change filled in.
Read and try
Books from the library that Experiments puts into practice.
-
Think Again Adam M. Grant Grant on thinking like a scientist about your own life: a hypothesis, a test, and a willingness to be wrong.
-
Designing Your Life Bill Burnett, Dave Evans Burnett and Evans on prototyping a change before you commit to it. An experiment is a prototype with an end date.
-
The 4-Hour Body Timothy Ferriss Ferriss's self-experiments, each one changed thing and a record of what happened.
Already have HourLife?
Experiments is under Reflection in the module list, and it is part of HourLife Premium. "+" starts an experiment, and the ⋯ menu holds "Insights", "Archived" and "Settings". Today's Evening mode carries an "EXPERIMENTS" card for every running experiment with a measure.
Experiments is in the Library, under Reflection. If it isn’t there, turn it on in Settings → Look & Feel → Modules. It needs HourLife Premium.
What would you like to know?
Ask a question about this feature or tell us what’s missing.
Try Experiments in HourLife.
HourLife is free to download. Experiments comes with HourLife Premium.
