The Surface Hypothesis: Why the Same Routine Runs Differently on Different Devices

Abstract

This paper proposes the Surface Hypothesis: adherence to a routine is shaped substantially by the surface the routine runs on — the physical medium and the attentional context in which execution happens. The same sequence of steps, run on different surfaces, produces meaningfully different completion rates.

The habit-tracking category has a well-documented 70–90% 30-day churn rate. Much of the research on why has focused on algorithmic and content-design variables: reminder cadence, gamification, motivational copy. We argue that an under-examined variable is surface itself — and that surface-aware design, applied in software and in hardware, may close a meaningful portion of the adherence gap.

Software alone can do a lot. A well-designed phone-based routine app is a significant improvement over what came before, and many of the techniques in this paper translate directly into app design. At the same time, a purpose-built physical surface — a small, quiet device dedicated to the ritual — addresses aspects of the adherence problem that software on a general-purpose device cannot fully solve. Both are worth building.

Framing the problem

A routine is a sequence of actions, in order, at a specific time. Whether the routine gets done depends on three things:

1. The user's intent. Do they want to run it today?
2. The design of the routine. Is it appropriately sized, specific, and cued?
3. The surface it runs on. What else is competing for the user's attention at the moment of execution?

The first two have received substantial research attention. The third has been largely treated as a constant ("people use phones") and therefore ignored as a design variable. We think that's a mistake.

What a surface does

For the purposes of this paper, a surface is the combined physical and digital environment in which a routine is executed. Its properties include:

• Physical presence. Is the surface visible without summoning it? A wall calendar has presence; an app icon does not.
• Attentional exclusivity. Does the surface do one thing, or many? A dedicated stopwatch does one thing; a phone does many.
• Environmental cues. Does the surface's appearance or placement signal when to use it? A running shoe by the door is a cue; a hidden reminder is not.
• Transition cost. How much effort does it take to go from "not running the routine" to "running the routine"?

Some of these properties can be influenced by software design (e.g. an app can reduce transition cost with a home-screen widget or a quick-start gesture). Others are inherent to the physical medium.

Phones as a surface

Phones are unusually powerful surfaces — they can hold entire libraries of routines, manage timers with millisecond precision, adapt prompts to context, and synchronize across devices. That power is why phone-based routine apps work at all, and why they have produced the best adherence rates the category has seen.

Phones also have real limitations as a routine surface. Specifically:

• The home screen is a compound environment, typically with 20–40 other applications visible or one tap away. Each of those applications is designed to capture attention if launched.
• Notifications arrive during routine execution by default, interrupting the attentional state the routine is trying to create.
• The same device is used for contexts (work email, social media, news) that are adversarial to the quiet attentional state most routines aim for.

None of these limitations are fatal. Well-designed software can mitigate each of them:

| Limitation | Software mitigation | |---|---| | Home-screen distraction | Quick-launch widget, one-tap start from alarm screen | | Notification interruption | Auto-enable Focus/DND for routine duration | | Adversarial device context | Wake-lock the screen to the current step, disable side navigation |

The Routine OS app incorporates most of these mitigations. They reduce the adherence gap significantly. They do not eliminate it.

Where software hits a ceiling

The residual gap — the adherence loss that remains even after software does everything it can — appears to correlate with two structural factors:

1. The user must still reach for the phone. The act of picking up the device, regardless of what happens next, puts the user in a posture that historically has included phone-checking behavior. Muscle memory is hard to override with software cues alone.

2. The device cannot credibly commit to a single role. Even with every notification silenced, every app hidden, and Focus mode enabled, the user knows the same device, seconds after the routine ends, will be used for scrolling. This latent context affects the attentional state during the routine.

Both of these are surface properties, not software properties. They suggest that incremental software improvements will produce diminishing returns past a certain point.

The case for a dedicated hardware surface

If the argument above is correct, there should be room for a surface that is physically aligned with the user's goal — a device designed to run routines and nothing else. Such a device would exhibit:

1. Single function. The device runs routines. Its physical form factor and firmware make it inappropriate for any other use.
2. Environmental presence. The device is visible in the space where the routine happens. Its presence is itself a cue.
3. Tactile control. Physical inputs, rather than a glass touchscreen that can summon other applications.
4. Firmware discipline. Features that would compromise single-function integrity (e.g. a feed, a notification tray, a browser) are structurally prohibited, not just disabled by default.

A device with these properties would not replace a phone-based routine app — the two serve different contexts and complement each other. A phone app is appropriate for flexible routines, on-the-go contexts, and data management. A dedicated hardware surface is appropriate for high-friction, attention- sensitive rituals — mornings, workday shutdowns, evening wind-downs.

Practical implications

For practitioners running routines today, with the software that exists, the Surface Hypothesis suggests several concrete moves:

• Place the phone face-down during routine execution. Even absent notifications, the visible screen is a cue to engage.
• Use Focus/Do Not Disturb modes to reduce cross-context interruption.
• Establish a consistent physical location for the routine — a specific chair, desk, or spot. Place supporting objects (water, journal, yoga mat) there the night before. Location and object cues do work the phone cannot.
• Treat the phone app as one component of the surface, not the whole surface. The environment around the phone matters as much as the app on it.

These changes are free and compatible with any routine-tracking software. Early data suggests they close a meaningful fraction of the adherence gap on their own.

Early prototype data

The remainder of this paper — including comparative adherence data between phone-only routines, phone-plus-environmental-design routines, and early prototype hardware trials, plus the rough design specification for the device we are building — is available as a PDF download below.

Get the full paper

Drop your email using the form below to receive the complete whitepaper as a PDF, along with occasional research updates. No drip sequence, no spam.