Habit Tracking

Habit tracking is the practice of recording whether you've completed a planned behavior each day. It can be as simple as marking an X on a calendar or as detailed as logging time, notes, and metrics in a dedicated app.

Tracking is not the habit. It is a measurement layer sitting on top of one, and like any measurement layer it changes the thing it measures — sometimes helpfully, sometimes not.

The two jobs a tracker does

Almost every argument about habit trackers comes from confusing these two jobs.

Job one: remembering. A tracker tells you what is due today. This is the job people notice, and it stops mattering once a behavior is automatic.

Job two: feedback. A tracker tells you what actually happened over the last eight weeks — which is almost never what you remember. This is the job that keeps paying off, and it is the reason to keep logs even for habits you no longer need reminding about.

A tracker that only does job one turns into a chore list. A tracker that does job two gives you something to make decisions with.

Where the practice comes from

Self-monitoring long predates apps. Benjamin Franklin kept a chart of thirteen virtues in a small book, focusing on one per week and marking a dot for every fault — an early habit tracker with a rotating focus rather than a total-coverage philosophy.

The "don't break the chain" method, usually attributed to Jerry Seinfeld marking calendar days for writing jokes, comes from a second-hand account by comedian Brad Isaac; Seinfeld has since said publicly he does not know where the idea got attached to him. Worth knowing, because the anecdote is often cited as if it were evidence.

The actual evidence for tracking is about progress monitoring generally. Harkin and colleagues published a meta-analysis in Psychological Bulletin in 2016 covering 138 studies and roughly 19,000 participants, and found that prompting people to monitor progress toward a goal produced a reliable improvement in goal attainment — with larger effects when the progress was recorded physically or reported publicly rather than just noticed. That is a finding about monitoring, not about streak counters, and the distinction matters for which metric you choose.

What each metric distorts

Metric	What it rewards	Where it misleads
Streak	Unbroken sequences	Treats one miss as total failure, so it peaks right before people quit
Completion rate	Overall volume	A 70% rate looks the same whether the misses were scattered or a two-week collapse
Bounce-back rate	Speed of return after a miss	Needs a few misses before it means anything; useless in week one

Read together they are informative. Read alone, a streak counter mostly measures how recently you got lucky — the case against it is laid out in why streaks don't work.

Six weeks of Tom's log

Tom tracked three habits from January onward: a 20-minute walk, 10 pages of reading, and no phone before 08:00.

By mid-February his completion rates were 71%, 88%, and 45%. The interesting part was not the percentages but the shape of the misses. Twelve of his seventeen walk misses were Tuesdays and Wednesdays — the two days he had late meetings that ran to 18:45. Reading misses were scattered singles, and he always came back the next day. The phone habit failed in one continuous eleven-day block starting the week he moved his charger to the bedside table.

Three different problems, and none of them were willpower:

• The walk needed rescheduling to 12:30 on two days a week, not more resolve.
• The reading habit was already healthy — scattered single misses with next-day recovery is what a working habit looks like.
• The phone habit needed the charger moved back to the hallway. It resumed at 79% once he did.

Without a log, all three read as "I'm being inconsistent lately."

Logging habits in Keel without it becoming a chore

The failure mode of tracking is overhead, so the useful setup is a small one.

• Track 2 to 5 habits, and only ones you are actively building. Behaviors that are already automatic can stay off the list; the log is for the ones still in question.
• Log from the Today screen once a day, ideally at a fixed moment — logging is itself a habit and benefits from an anchor.
• Use "skipped" for a deliberate rest day and leave a miss as a miss. Marking a planned rest day as skipped keeps it out of your completion rate; marking a genuine miss as skipped just hides the pattern you need to see.
• Write one line in the note when you miss. "Meeting ran late," "away from home," "didn't want to." Three weeks of these is where the actual diagnosis comes from, and it is the part no metric can compute for you.
• Review weekly, not daily. Daily numbers are noise. Look at which weekday the misses cluster on and whether you returned within 24 hours.
• Link each tracked habit to a goal, so a weekly review can ask whether the habit is moving anything, not just whether you did it.

If you are still choosing a tool, the tradeoffs between paper, spreadsheets and apps are covered in our roundup of free habit tracker apps.

When to stop tracking

Retiring a habit from the tracker is a legitimate outcome, not quitting. Reasonable signals: you have done it without checking the app for two or three weeks, missing it feels odd, and it no longer appears in your weekly review as a question. Log it out and give the attention to the next habit that still needs job one.

Related terms

• Self-monitoring — the clinical term for the same practice, usually applied to symptoms, food or mood rather than intended habits.
• Streak — one specific metric a tracker can display, not a synonym for tracking.
• Habit stacking — a technique for building the habit; tracking is how you find out whether it took.
• Quantified self — broad-spectrum personal data collection, often passive and sensor-driven; habit tracking is deliberate, manual and limited to behaviors you chose in advance.