What your completion rate actually means

Completion rate is the most-quoted number in habit tracking and the most frequently misread. A 60% month can mean you are doing well, or that you are quietly failing, and the number alone cannot tell you which.

The reason is that a completion rate is a ratio, and almost nobody examines the denominator. Change what you scheduled and the same behavior produces a wildly different percentage. Until you know how the denominator was built, the numerator is not interpretable.

What the number actually is

Completion rate is completions divided by opportunities over some window.

Every part of that is a design decision. Which tasks counted as scheduled on a given day. Whether a three-times-weekly habit contributes one opportunity or three. Whether all tasks count equally. Whether the window is calendar-aligned or rolling. Two apps can report 55% and 80% for identical behavior, and both can be correct under their own definitions.

In SELV, the denominator is the sum of the weights of the tasks actually active on that day, and the numerator is the sum of the weights you completed. Two consequences follow. First, a task that was not scheduled today does not count against you — a three-times-weekly habit creates three opportunities a week, not seven. Second, because weights run 1 to 5, skipping the hard thing costs more than skipping the easy thing.

That second property is the one that makes the number honest. Under flat scoring, you can complete three trivial tasks, skip the one that mattered, and post 75%. Under weighted scoring, the same day reports something much lower, because the measure now tracks where your effort actually went. What XP should actually measure covers the weighting mechanics.

The four ways it lies

1. An inflated denominator

The most common distortion, and it is self-inflicted at setup.

If you set a gym habit to daily when you realistically train three times a week, you have manufactured four guaranteed failures every week. Your completion rate drops to around 43% for that habit, and your overall number drops with it. Nothing about your behavior is wrong. The schedule is a lie, and the percentage is faithfully reporting the lie.

This is the single highest-leverage fix available to most people. Set the cadence you will actually hit. The same three gym sessions at a three-times-weekly cadence report 100%, and the data becomes usable.

2. A deflated denominator

The opposite failure, and harder to notice because it feels good.

A list of five trivial tasks will report 95% indefinitely while nothing in your life changes. The number is accurate and meaningless. If your completion rate has been above 90% for two months and you cannot name anything that has improved, your tasks are too easy — which is a reason to raise the difficulty, not to congratulate yourself.

A useful heuristic: a well-calibrated habit set runs somewhere in the 70–85% band. Consistently above 90% suggests you have room to add difficulty. Consistently below 60% suggests the schedule or the quantity is wrong.

3. The average hides the shape

Two people both at 70% over a month:

The first completed roughly 70% every week. The second ran at 100% for two weeks and 40% for two weeks. Same average, completely different situations. The first has a stable system slightly below capacity. The second has a novelty crash, and the trend is heading somewhere bad.

An average over a long window destroys exactly the information you need. Look at the weekly series, not the monthly mean. Reading your habit heatmap covers how to see the shape at a glance.

4. The overall number hides the distribution

A 70% month where every category sits near 70% is a balanced system. A 70% month made of Vitality at 95% and Nourish at 25% is something else entirely — one area is working and one has effectively stopped.

The aggregate is arithmetic, not insight. The per-category breakdown is where decisions come from, which is most of the reason for having categories at all. In SELV that is five: Vitality, Discipline, Nourish, Order and Self-care.

This also works in your favour on bad weeks. A 38% week looks like total failure as a single number. Split by category it frequently shows the important things held and the optional things dropped, which is competent triage rather than collapse.

What a good rate looks like

There is no universal target, and anyone quoting one is guessing. What matters is the relationship between the number and your own configuration.

The 40–55% band is where most abandoned trackers live, and it is almost always a configuration problem rather than a discipline problem. How many habits should you track at once covers the fix.

The three questions worth asking of the number

Rather than asking whether your rate is good, ask:

Is it trending? Direction over six to eight weeks carries far more information than level. 62% rising from 50% is a system coming together. 62% falling from 80% is a system coming apart, and the two demand opposite responses.

Which habit is responsible? Overall rates are almost always dominated by one or two outliers. A single habit at 20% can pull a whole month down several points. Find it, then decide: fix the cadence, shrink the behavior, or drop it.

Which day is responsible? Weekday breakdowns are the most underrated view in habit analytics. Most people have one structurally bad day, and almost nobody knows which one until they look. Once you know, you can fix it with a schedule change rather than with resolve.

Completion rate versus cumulative progress

These answer different questions and a serious tracker should report both.

Completion rate is a window measure. It tells you how you are doing lately, and it resets with every window — it does not care what you did in March.

Cumulative XP is a total. It tells you what you have built, it never goes down, and it is deliberately insensitive to a bad fortnight.

You need both because they fail in opposite directions. The rate alone makes a bad month feel like the whole story. The total alone lets you coast on past accumulation while your current performance quietly degrades. SELV keeps rolling averages for the first question and permanent XP for the second, and shows them side by side rather than collapsing them into a single score.

How to actually use it

A short routine that takes about five minutes:

1. 01Look at the weekly series, not the monthly average. Six to eight points. You are looking for direction and volatility, not a level.
2. 02Then the per-category breakdown. Anything more than about 25 points below your overall rate is the thing to work on.
3. 03Then the per-habit rates. Identify the single worst performer and decide its fate: fix cadence, shrink, or cut.
4. 04Then the weekday view. If one day is consistently 20+ points below the others, that is a scheduling problem with a scheduling solution.
5. 05Change one thing. Not four. If you change four, you will not know which one worked.

Do this every two weeks. Daily data is noise — a single day tells you nothing, and checking it daily mostly trains anxiety. The weekly review covers the routine in more detail.

One number people should stop quoting

A note on comparing your rate with anyone else's: it is meaningless, and more so than most metrics.

Your denominator was constructed by you. Someone reporting 95% may have five trivial daily habits; someone reporting 60% may have five hard ones at honest cadences and be doing considerably more actual work. Weighted and unweighted systems are not comparable at all, and neither are two unweighted systems with different task difficulty.

The only legitimate comparison is against your own history, with the configuration held constant. Change the list and you have reset the baseline — which is a good reason to note the date whenever you make a significant configuration change.

The honest part

A completion rate is a measurement, not a grade. It describes a configuration meeting a life, and when the number is bad the configuration is usually the thing that is wrong.

That is why honest reporting matters more than encouraging reporting. If you hit 23%, the number should say 23% — not round up, not soften, not congratulate you on showing up. A softened number removes the only signal you had that something needs changing, which is a strange thing for a measurement tool to do.

Further reading: the complete guide to gamified habit tracking, reading your habit heatmap for the visual version, and habit tracking burnout for what happens when the number starts running your mood.