A delivered message is not an opportunity
At 7:00 p.m., the phone displays “Read.” You are on a train, the book is at home, and dinner is not decided. The notification worked technically. The plan did not work behaviourally.
This distinction can interrupt a common escalation: one ignored reminder becomes two; two become an earlier warning plus a final warning; soon the system produces a cloud of prompts that still point to no usable moment.
Digital behaviour-change research often uses fixed time prompts because time and notification delivery are easy for devices to track. A 2024 systematic review of digital physical-activity interventions noted that fixed time cues dominate the literature, while low response to push notifications and limited adaptation to context remain challenges. The review does not prove that notifications inevitably lose power. It does support being cautious about treating delivery as equivalent to action.
The useful question is: What job is this reminder meant to do?
- Remembering: bring a decided action to mind at a usable moment.
- Preparing: tell you to put materials in place before a later opportunity.
- Checking: invite a brief review after the opportunity has passed.
One alert should not silently perform all three jobs.
The five-link reminder audit
Look at the last three occasions when the reminder did not lead to the action. Classify the earliest broken link.
1. Delivery: Did the prompt arrive where you could receive it?
The phone was on silent, the app lacked permission, the alert appeared on a device left elsewhere, or the schedule did not match the correct days.
Adjustment: Fix the channel or schedule once. Do not redesign the behaviour to solve a settings problem.
2. Notice: Did the prompt enter your attention?
It may have arrived among many unrelated alerts, while you were presenting, driving, sleeping or engaged in another task. “I saw it later” is different from “I chose not to act.”
Adjustment: Remove competing low-value alerts, change the channel, or use a persistent environmental cue at the place of action. Do not make claims about your motivation from a prompt you never meaningfully noticed.
3. Meaning: Did the message name a clear first move?
“Exercise,” “focus” or “be productive” requires planning. Even “read” leaves the material, location and finish undefined.
Adjustment: Put the action in the plan, not necessarily in the notification: “After I place my tea on the desk, open the book to the marked page and read one paragraph.” The reminder can then say only enough to point back to that plan.
4. Fit: Could the action reasonably happen then?
The prompt may consistently collide with work calls, care responsibilities, travel or a room where the materials are absent.
Adjustment: Move the opportunity, make a constrained version, or tie the action to an event that occurs closer to the usable moment. The answer is not automatically “use an event cue”—a randomized study of nutrition habits found no group-level difference between routine-based and time-based cue plans. Repeatedly enacting the plan in response to the chosen cue was the stronger predictor. Choose the cue you can actually encounter and use.
5. Return: After dismissal, was there a defined next opportunity?
Many reminders offer only “now.” If now does not fit, dismissal can feel like failure or push the task into a vague later.
Adjustment: Add a return rule: “If the evening cue is blocked, the next opportunity is after tomorrow’s first tea—not a late make-up session.” For a non-daily habit, name the next genuine weekly opportunity rather than saying “tomorrow.”
Decide whether the reminder is the cue or scaffolding
These are different designs.
The reminder is the cue
The alert itself is the intended signal. This can work when the time is reliably usable, the action is immediately available, and the device is normally present. The plan might be: “When the 3:00 p.m. alert appears during my break, I will write tomorrow’s first task on the card beside my keyboard.”
If the time shifts daily or the action cannot begin from the phone, the notification is a poor sole cue.
The reminder supports a real-world cue
The actual cue is an event—finishing lunch, closing the laptop, arriving home—and the notification temporarily helps you remember the connection. In that design, the alert should eventually become redundant if the surrounding situation begins prompting the action.
You can test that by delaying or removing one reminder for two weeks while leaving the event and materials unchanged. If the event still brings the action to mind, the scaffolding may have done its job. If the action disappears, restore the support and inspect whether the event is noticeable and stable enough.
Do not remove every support at once. A clean test changes one thing.
Rewrite the system, not just the message
Use this template:
Example:
Now the alert does not command an abstract behaviour at an arbitrary time. It prepares a visible environment for a later cue.
Another example for a variable schedule:
The fallback is part of the design, not an admission that the main plan failed.
When to reduce, retain or replace a reminder
Reduce it when the real cue regularly brings the action to mind before the alert, the reminder adds no new information, and a two-week test does not make starting harder.
Retain it when it reliably appears at a workable moment and solves a genuine remembering problem with little extra noise.
Replace it when the time repeatedly conflicts with the opportunity, the message requires fresh planning, or the device is not present where the action happens.
Repurpose it when preparation is more valuable than prompting: set out the book, fill the bottle for later, put the document in the start-up folder.
End it when the behaviour no longer matters. An alert cannot repair a goal you no longer choose.
Run a seven-opportunity experiment
For the next seven real opportunities:
- Keep one primary cue.
- Give one reminder one explicit job.
- Define the smallest honest version of the action.
- Record the earliest broken link: delivery, notice, meaning, fit or return.
- Adjust only that link for the following opportunity.
At the end, keep the fewest prompts that make the next decision clear. The aim is not a silent phone at all costs. It is a system in which every remaining reminder earns its interruption.
Evidence notes and sources
- Keller, J. et al. (2021). Habit formation following routine-based versus time-based cue planning: A randomized controlled trial. British Journal of Health Psychology, 26, 807–824. Both cue-planning groups increased in automaticity; no between-group difference was found, and plan enactment predicted automaticity. This supports testing cue fit rather than declaring one cue type universally superior.
- Stojanovic, M., Grund, A., & Fries, S. (2022). Context stability in habit building increases automaticity and goal attainment. Frontiers in Psychology, 13, 883795. More stable contexts predicted greater automaticity in two longitudinal datasets; this does not mean fixed clock time is always the relevant context.
- Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement. Advances in Experimental Social Psychology, 38, 69–119. Supports specifying the situation and response in an if–then plan; it does not show that adding more notifications creates stronger habits.
- Zhu, Y. et al. (2024). Digital Behavior Change Intervention Designs for Habit Formation: Systematic Review. Journal of Medical Internet Research, 26, e54375. This review of digital physical-activity interventions identifies the dominance and limitations of fixed time-based prompts; its design synthesis is not evidence that every user habituates to alerts.