AI Insights
Your habits are
talking to each other.
Traction's AI engine analyses every entry you log to surface five distinct types of meaningful pattern — connections you'd never notice just by looking at a calendar.
Same Day
72%
correlation strength
Odds Ratio
47%
more likely
Time-Delayed
68%
peaks at 2 days
Pattern
3.2×
more likely · after 3 days
Habit Cycle
4×
per week
7 weeks
active
3 weeks
break
How it works
Most apps just show you what you did. Traction tells you what it means.
Every night, Traction's analysis engine runs across your entire history — every workout, sleep score, mood rating, and custom metric you've ever logged. It looks for five statistically significant pattern types that reveal how your habits genuinely influence each other.
The result lands in your Insights feed: a personalised dashboard of discoveries ranked by statistical strength. No noise, no generic advice — just findings that are specific to you.
When two habits move together
Some habits are secretly in sync. Same-Day Correlation detects when two of your tracked metrics consistently rise or fall on the same day — a signal that they're influencing each other in real time.
- Positive correlation: both habits tend to be high on the same days.
- Negative correlation: one habit tends to be low when the other is high.
- Strength is measured from 0–100% — higher means the connection is clearer and more consistent.
- Only shown when the pattern is statistically robust across your full history.
Example
"On days you log a Morning Run, your Healthy Eating score is 72% correlated — they almost always occur together."
Same Day
72%
correlation strength
How much more (or less) likely?
Correlation tells you two things move together. Odds Ratio tells you by how much. This insight calculates the probability of a positive outcome on days when a specific habit is logged versus days when it isn't.
- "More likely" insights: logging habit A makes outcome B significantly more probable.
- "Less likely" insights: certain habits reduce the probability of a given outcome.
- Expressed as a percentage uplift — e.g. "47% more likely to have a productive day".
- Useful for building intentional morning or pre-event routines.
Example
"On days you meditate, you are 47% more likely to rate your day as Productive."
Odds Ratio
47%
more likely
The effects that show up tomorrow
Not all cause-and-effect relationships are immediate. Lag Correlation scans across a range of time windows — 1 to 7+ days — to find when the connection between two habits peaks. Some habits reward you a day later. Some take a week.
- Scans lag windows from 1 to 7 days to find the peak correlation point.
- The sparkline shows how correlation strength rises and falls across each delay.
- Highlights the single day lag with the strongest statistical connection.
- Useful for planning recovery, nutrition timing, or pre-event preparation.
Example
"Strength Training correlates most strongly with Deep Sleep quality after a 2-day lag — your body’s recovery peaks on night two."
Time-Delayed
68%
peaks at 2 days
Consistency is the multiplier
A single good day doesn't always move the needle — but a sustained streak does. Pattern Effect looks at rolling windows of habit consistency and finds the streak length that most strongly predicts a positive outcome.
- Analyses rolling windows of 1–7 days of consistent habit logging.
- The bar chart shows how predictive power changes with streak length.
- The peak bar shows the window where the habit’s effect is strongest.
- The multiplier shows how much more likely the outcome becomes at that window.
Example
"3 days of consecutive Meditation makes a Peak Focus day 3.2× more likely than your baseline."
Pattern
3.2×
more likely · after 3 days
The rhythm underneath your routine
Some habits aren't daily — they pulse. Habit Cycle analysis detects the natural on/off rhythm of a given behaviour: how often it occurs in active phases, how long those phases last, and what break length typically follows.
- Measures frequency (times per week) across your logged history.
- Identifies the typical length of active phases and rest phases.
- The ring shows the proportion of active time to total time.
- Useful for designing training blocks, rest weeks, or cyclical routines.
Example
"Your Gym habit runs in 7-week active blocks at 4 sessions per week, followed by a natural 3-week deload."
Habit Cycle
4×
per week
7 weeks
active
3 weeks
break
Why this matters
Data you can act on, not just read.
Most self-improvement advice is generic. Traction's insights are built from your data, your history, your patterns. They answer the questions that actually matter: What should I do today to feel better tomorrow? Which habits are actually moving the needle?
Built from your history
Every insight requires a minimum number of data points to reach statistical significance. No guesswork — only findings your own numbers support.
New insights every week
As you log more, the engine runs again. Patterns get stronger, new connections emerge, and old ones are retired when your habits change.
Five different lenses
Different insight types reveal different truths. A same-day correlation and a lag insight about the same habit pair can both be true — and both useful.
Get started
Start logging. Start discovering.
The first three months are completely free. No credit card, no commitment. Just your habits and the patterns hiding inside them.