The two categories, stated precisely
Habit trackers store behaviours and whether they occurred. The core interaction is a checkbox, once per day per habit. The output is a grid, a streak, and some statistics. Examples of the shape: a daily list of habits, a monthly heat map, a longest-streak number.
Routine apps store ordered sequences and run them. The core interaction is starting a routine and being carried through it step by step. The output is that the routine happened.
These sound adjacent. In practice they solve different problems, and only one of them is the problem ADHD actually has.
What a habit tracker assumes about you
A tracker is built on an implicit theory: that the reason you did not do the thing is that you forgot, or you were not sufficiently motivated, and that awareness plus a streak will fix both.
For a lot of people that theory holds. If your executive function is intact and the barrier is genuinely just attention and follow-through, a visible streak provides real motivational pull and a grid provides real feedback.
For ADHD, the theory is usually wrong in a specific way. You did not forget. You knew about it all day. You thought about it, felt bad about it, and did not start. Awareness was never the bottleneck; initiation was. A tracker offers no help with initiation whatsoever — it engages only after the behaviour is over.
Worse, it adds a step. Now there is the thing, and there is also reporting the thing. On a good day that costs five seconds. On a bad day the reporting is one more small obligation you did not meet, and it accumulates its own guilt.
What a routine app assumes about you
A routine app is built on a different theory: that you know what to do, and the failure is in the execution machinery — starting, ordering, switching, estimating time.
So it takes those over. It holds the sequence, so working memory does not have to. It shows one step at a time, so there is nothing to decide mid-routine. It times each step externally, so time blindness has less to distort. It carries the current step onto the Lock Screen, so a distraction costs thirty seconds instead of the whole routine.
Tracking still happens, but as a by-product of running the routine rather than as a separate reporting duty.
Side by side
| Habit tracker | Routine app | |
|---|---|---|
| Core object | A behaviour | An ordered sequence |
| When it engages | After the fact | During |
| Main interaction | Tick a box | Start, then follow |
| Helps with initiation | No | Yes — one tap to running |
| Helps with sequencing | No | Yes, that is the point |
| Helps with transitions | No | Only if it runs outside the app |
| Handles time blindness | No | Yes, if timing is visible |
| Adds a daily obligation | Yes, reporting | No |
| Typical failure | Abandoned after a broken streak | Abandoned if setup is heavy |
Where routine apps fail
This is not a clean win, and it is worth being honest about the failure modes.
Setup cost. A routine app needs the sequence defined before it does anything. If building the first routine takes an evening, it will not happen, and the app dies before it is ever used. Templates are the only real answer to this, and an app without good ones is asking for work you cannot reliably supply.
Rigidity. A fixed sequence can become another thing to fail at, especially if the app treats deviation as an error. Good routine apps let you skip a step, reorder on the fly, and finish a routine partially without any of it being marked as failure.
Notification fatigue. Anything that pings you into a routine will eventually be ignored, and then muted, and then the app is dead weight. Ambient presence — a widget, a Lock Screen activity — outperforms alerts because it is available when you look rather than demanding you look.
Where trackers still earn their place
Two cases:
- Discrete behaviours with no sequence. Taking a supplement, flossing, a single daily walk. There is no routine to run; there is just a thing, and a record is genuinely all you need.
- Diagnostic periods. Two weeks of tracking to find out when your energy actually peaks, or whether a change helped, is legitimately useful. It is a temporary instrument, not a permanent habit.
The mistake is treating a temporary instrument as a permanent system.
The streak problem applies to both
Both categories tend to ship streak counters, and streak counters as usually implemented are actively harmful for ADHD. A number representing months of effort, deleted because of one missed Tuesday, produces exactly the "I always ruin things" conclusion that keeps people out of the app permanently.
Whichever category you choose, the thing to actually check is what the app does on a missed day. Freezes — protected days granted as the streak grows and spent automatically — are the standard fix and they change the emotional character of the whole thing. This matters more than any other feature comparison in this article.
How to choose
Ask what actually happens when the behaviour does not happen.
- If the answer is "I forgot" → a tracker may genuinely help.
- If the answer is "I knew, and I did not start" → a tracker will not help, and a routine app might.
- If the answer is "I started and then lost the thread halfway" → you need a routine app that runs outside itself, on a Lock Screen or a widget.
- If the answer is "I did it for eleven days, missed one, and never opened it again" → your problem is the streak design, and you should choose on that basis alone.
For most people with ADHD the honest answer is the second or third one, which is why the tracker shelf is full of apps that were used for nine days.
Where Orbit sits
Orbit is a routine app, not a tracker. Routines run as an ambient orb, one step at a time, continuing on the Lock Screen and in the Dynamic Island so a distraction does not end the sequence. Templates mean the first routine takes under a minute instead of an evening. Streaks exist, but they earn freezes that are spent automatically on missed days, so the record survives a bad week. There is no daily reporting to do, because running the routine is the record.