Why routines die
“Morning routine” is a stack of sequential actions — wake, teeth, clothes, bag, out — and each transition is an initiation event with a cue attached. For ADHD the cue is the fragile part: the internal “next, do X” signal drops, the sequence stalls, and by 9am the day is already running late. A routine app’s real job is cue externalisation: show the next step, on schedule, without requiring you to remember it. This is the same principle as every externalised system — applied to sequences instead of lists.
What the good ones do
- Visual step sequences: a card-per-step flow you can follow without holding the plan in memory.
- Time-based and event-based triggers: “at 7:40” or “after the kettle” — anchors that fire without your attention (time blindness mean time-based triggers need redundancy).
- No streak shame: the chain-mechanic apps create weekend guilt that kills the whole routine (see the planner guide on guilt on miss).
- Short interaction cost: a one-tap “done” beats a checklist you must open and maintain.
The honest landscape
Dedicated routine/habit apps (step-sequence class)
Purpose-built for sequences: visual steps, timing, gentle reminders. They externalise the cue well; watch for notification fatigue — the app only works if its timing matches when you actually need the cue.
Visual timer + habit stack (physical timers)
The lowest-tech option: a wedge timer per block and a printed step list. Zero apps, zero notifications, fully visible — the visual timers guide has the hardware.
Daily-plan apps with routine slots (Sunsama/Tiimo class)
Routines live inside the day plan — time-shaped and visible alongside everything else. Strong when the day varies; weaker when the sequence needs its own pacing.
Proactive AI planners
The emerging end: routines triggered at the right moment, resurfaced when missed, broken down when needed — the honest scorecard is AI ADHD planners.
Building your first routine with a tool
- Choose the one routine that matters most (morning is the highest-leverage start) and write it as 5–7 physical steps.
- Set the first cue externally — alarm at the entry point, step cards visible, timer per block.
- Plan for the average bad day, not the perfect day: a 3-step minimum version that counts as “done”.
- Iterate weekly by what actually fell off — and remember a fallen routine is data, not a verdict (burnout is the thing to watch if nothing sticks).
Routines, daily plans, and timers are three layers of the same externalisation principle. The directory shows how they fit together without stacking ten tools at once.