How to Define a Habit You Can Track Daily
Vague goals are untrackable by design. Use the two-observer test, a rewrite table, and six measurement types to define a habit you can log daily.
Open any habit tracker and it asks you the same question every night: did you do it, yes or no? That question is easy to answer if your habit is “read ten pages before lights out.” It is impossible to answer if your habit is “read more.”
Willpower usually gets the blame for an abandoned habit tracker, but the real problem sits further upstream. The habit was never defined well enough to produce an answer, so the log filled up with guesses, and guesses are not worth opening an app for.
Fixing that takes about five minutes and one piece of paper.
”Exercise more” is an intention, not a habit
An intention describes a direction. A habit describes an action, and trackers only work on the second kind.
Try the two-observer test on your current habit. Imagine two reasonable people who watched your entire day, and ask whether they could disagree about whether the habit counted. If they could, the habit is not defined yet.
“Exercise more” fails immediately. Does carrying groceries up three flights count? A ten-minute walk to the bus? Observer A says yes, observer B says no, and now you get to decide, at 11pm, when you are tired and would quite like to say yes.
That last part is the real cost. Ambiguity does more than make logging annoying: it hands you a renegotiation window at the exact hour when your motivation is lowest and your reasons for dropping the bar sound most convincing. A vague habit invites you to lower the standard every evening.
Locke and Latham spent 35 years testing this. Their conclusion, drawn across hundreds of studies, is that specific goals outperform vague or “do your best” goals. The reason is fairly plain. A specific goal tells you when you are finished, and a vague one never does.
The three questions that make a habit trackable
You can run this in sixty seconds.
1. What exactly do I do? A verb and an object. “Walk,” “read,” “stretch,” “cook.” It has to be something you do rather than something you avoid. Gardner, Lally and Wardle (2012) are direct about it: it is not possible to form a habit for not doing something. Their advice is to select a new behaviour rather than simply give up an existing one.
That matters if your habit is “no sugar” or “no phone after 9pm.” Keep the avoidance as your rule if you like, but the thing you log should be the substitute: sparkling water with dinner, phone on the charger in the kitchen, a book on the nightstand. Log the action that displaces the thing you are avoiding.
2. How much counts? A number, or an explicit “any amount.” Both are valid definitions. Leaving it open is not, because open is what you argue with later.
3. By when? A daily window. Clock times work, but cues work better. Gardner and colleagues note that the cue must be salient enough in daily life that you encounter and detect it frequently and consistently. “After I brush my teeth” fires reliably. “In the evening” does not.
Run those three questions over your habit and the rewrite tends to be obvious:
| Vague intention | Trackable definition |
|---|---|
| Exercise more | 20 minutes of deliberate movement before dinner |
| Read more | 10 pages before lights out |
| Drink more water | 8 glasses by 9pm |
| Be more present | One 10-minute phone-free walk after lunch |
| Eat healthier | 30g of protein at breakfast |
| Stop doomscrolling | Phone on the kitchen charger by 9:30pm |
Every rewrite on the right has a verb, an amount, and a window, and each one answers yes or no without a debate.
Define the behaviour, not the outcome
Of everything on this page, this is the distinction people skip most often, and skipping it is what ends the run.
Weight, body fat, mood, income and resting heart rate are outcomes. They lag, they bounce around for reasons unrelated to what you did, and you do not directly control them. Walking 8,000 steps, cooking dinner, and doing 20 push-ups are behaviours. You decide those.
Behaviour change researchers treat these as formally separate techniques. The BCT taxonomy (Michie et al., 2013) lists “goal setting (behaviour)” and “goal setting (outcome)” as distinct entries, with action planning as a third. Swapping one for the other breaks the log, because a daily tracker can only honestly hold the first.
The practical rule: track the behaviour daily, measure the outcome weekly.
One exception trips people up. Stepping on the scale is itself a behaviour. So is taking a progress photo. The act of doing it is what you tick off, and the number or image it produces is the outcome. Confuse the two and you will feel like you failed on a morning where you did exactly what you planned and the scale went the wrong way.
In LockIn30, that split is built into the setup: a measurement task logs a number over time, and a photo task logs the shot itself. Neither one grades you on what the number says.
Pick the measurement type that matches the habit
“Make it measurable” is where most guides stop. The more useful question is what kind of measurable, because the wrong unit quietly corrupts the log.
| Type | Use when | Sounds like | Failure mode |
|---|---|---|---|
| Checkbox | The act is all or nothing | ”Took vitamins” | Hides big swings in effort |
| Counter | Volume is the point | ”8 glasses of water” | Target set too high on day one |
| Duration | Time is the honest unit | ”20 minutes of reading” | Encourages clock-watching |
| Photo | Proof or visible change matters | ”Front-facing progress photo” | Lighting and framing drift |
| Text | Content matters more than count | ”Three lines of journaling” | No floor, so entries shrink to nothing |
| Measurement | You are logging a number over time | ”Morning weight in kg” | Outcome gets mistaken for behaviour |
Choosing the wrong type is a definition error wearing a costume. “Meditate” as a checkbox invites a 40-second session that technically counts. As a duration (“10 minutes”), the argument never starts.
The reverse mistake is just as common. Making “took vitamins” a duration is absurd, and making “journal” a counter produces three-word entries that hit the target and mean nothing.
Photo habits deserve a specific warning. The photo itself is trivially easy to log, which is why the discipline has to live in the setup rather than the target: same spot, same light, same time of day. We covered that in detail in the progress photo guide, and the short version is that consistency of conditions is what you are actually buying.
Write the bad-day floor before day one
Every good habit definition has two lines, not one. The normal version, and the version that still counts on your worst day.
BJ Fogg built an entire method around shrinking behaviours until they survive low motivation, and his recipe format (“after I [anchor], I will [tiny behaviour]”) is worth stealing. For a fixed-length challenge, the useful adaptation is a defined minimum decided in advance, in writing, while you are calm, rather than improvised at 11pm, when improvising means zero.
Reading 10 pages has a floor of one page. A 20-minute run drops to a walk around the block, and journaling three lines drops to one line. Calling that cheating misses what the floor is for. It keeps the chain intact through a fever, a delayed flight, or a genuinely terrible Tuesday.
While you are there, pre-answer the four edge cases that will otherwise ambush you:
- Travel days. Does a hotel room count? Decide now.
- Sick days. Does the floor apply, or is illness an explicit exemption?
- Double sessions. Do they bank for tomorrow? The answer is no. Habits are about repetition in context, and doing two today does not repeat anything tomorrow.
- Half-completions. Six pages instead of ten. Either it clears the floor or it does not, and that is a question the floor already answered.
Predefining the floor is also what makes strict tracking survivable. Strict mode with no floor is a coin flip. Give it a floor and it turns into a commitment you can actually keep, which is a large part of how to choose between strict and forgiving tracking.
Turn the definition into an if-then plan
The last step is the one with the strongest evidence behind it. Take your definition and attach it to a cue in the form: if [situation], then I will [behaviour].
Gollwitzer and Sheeran’s meta-analysis of 94 independent tests, covering more than 8,000 participants, found a medium-to-large effect (d = .65) of implementation intentions on goal attainment. Sheeran, Listrom and Gollwitzer revisited the question in 2024 across 642 tests and found the effects hold, with the contingent if-then format and rehearsal doing much of the work.
The cleanest real-world demonstration comes from Milkman and colleagues (2011), who mailed flu vaccination reminders to employees. From a base rate of 33.1%, a prompt asking people to write down a date lifted vaccination by 1.5 points, which was not statistically significant. A prompt asking for a date and a time lifted it by 4.2 points, which was. Same reminder, same clinic, same people. The specificity was the active ingredient.
Which gives you the finished template:
After [existing cue], I will [verb + object + amount]. If that is impossible, I will [floor]. I will log it as [checkbox / counter / duration / photo / text / measurement].
Worked example: After I make coffee, I will read 10 pages. If that is impossible, I will read 1 page. I will log it as a counter with a target of 10 pages.
One timeline note before you start. Lally et al. (2010) tracked 96 people forming daily habits and found a median of 66 days to automaticity, with a range from 18 to 254 days. A 30-day challenge is not long enough to make most habits automatic, and it is not supposed to be. What it does buy you is enough time to prove the definition holds up and to build a streak worth protecting. The same study found that missing one opportunity did not materially affect habit formation, which is a useful thing to know on day nine.
The definition is the setup screen. One habit, one verb, one amount, one window, one floor, one input type, one duration. If you can write those seven things on a line, you can answer yes or no tonight, and every night after.
If you want to sketch the run before committing, the 30-day habit challenge tracker gives you a printable grid, and the end date calculator tells you what day 30 lands on.
Then pick your habit, set the duration, and log it once a day. Download LockIn30 if you want the definition, the streak and the photos to live in one place on your phone, with no account and nothing leaving the device.
Frequently Asked Questions
How do I know if my habit is specific enough to track?
Run the two-observer test. If someone who watched your entire day could reasonably disagree with you about whether the habit counted, it is not defined yet. Add three things: a verb and an object, an amount, and a daily window. If all three are present, the yes or no answer settles itself.
What is a binary habit?
A habit with a clean yes or no answer at the end of the day, like took vitamins or read ten pages. Binary habits are the easiest to log because the completion rule was settled in advance. Counters and durations are for habits where volume or time is the actual point.
Can I track a habit that is about not doing something?
Not directly. Gardner, Lally and Wardle (2012) put it plainly: it is not possible to form a habit for not doing something. Define the replacement behaviour and log that instead. Sparkling water instead of the second beer, a book on the nightstand instead of the phone. The avoidance can still be your streak rule.
Should I track the habit or the result?
Track the behaviour daily and check the result weekly. Weight, mood, body composition and revenue are outcomes you influence but do not decide. The behaviour is the only thing you can honestly tick off tonight, and it is the only thing a daily log can measure without lying to you.
How specific is too specific?
When the definition becomes fragile. If a five-minute delay, a different room, or a missing piece of equipment means it does not count, you have written a rule that only works on perfect days. Be specific about what counts and flexible about where and how it happens.
What is an implementation intention?
An if-then plan that links a cue to a behaviour: if it is 7am and I have poured coffee, then I will read ten pages. Gollwitzer and Sheeran's meta-analysis of 94 tests found a medium-to-large effect on goal attainment, and later work across 642 tests confirms the contingent if-then format is doing the work.
How many habits should I define at once?
One, for a 30-day run. Every extra habit multiplies the daily decisions and the number of ways the log can go ambiguous. One well-defined habit that survives 30 days beats three fuzzy ones that all quietly stop getting logged in week two.
How long before the habit feels automatic?
Lally et al. (2010) found a median of 66 days, with individual participants ranging from 18 to 254 days. A 30-day challenge is long enough to build the streak and prove the definition works. Automaticity usually shows up after the challenge ends, not during it.