# Why I built Pause Moment: the personalization stack that finally worked

Last updated: 2026-05-22

> The lived-experience story behind Pause Moment. Why every reminder app failed me, and the personalization stack — photo, words, sound, timer, silent cue, protect-my-pause, safe-people — that finally worked. Written by Ezra Halevi, founder.

# Why I built Pause Moment: the personalization stack that finally worked

By Ezra Halevi, founder of Pause Moment. Built in Kotlin. This is a founder-story piece. I wrote it. The research below is what convinced me the problem was structural, not personal.

**TL;DR.** ADHD medication adherence drops to 20 to 40 percent within 12 months. Generic notifications habituate within two weeks. The fix is not louder alarms. It is a stack of personal cues that your attention does not learn to filter out, held open by a screen lock you cannot dismiss with one thumb.

*About this article: A founder-story piece synthesizing 9 peer-reviewed and clinical-reference sources on ADHD medication adherence, notification habituation, behavioral salience, and design friction. Lived-experience reflections on the apps I tried, not a feature comparison. For the comparison version, see the listicle linked at the end. Not medical advice.*

A few years ago I started missing my own daily medication. I have ADHD. I also have a separate daily mental-health prescription that I cannot afford to miss. Every reminder app I tried failed me in the same specific way. The alarm would fire. I would dismiss it. I would not remember dismissing it. The dose would be missed. The next day, the same.

I am an Android developer. So I did the thing developers do when their tools fail them. I read the documentation for every medication-reminder app in the top of Google Play, bought the premium tiers, and noted what each one was actually doing under the hood. The conclusion took me by surprise. Almost every app was building toward the same wrong axis: the assumption that the dismiss-the-reminder failure was a failure of intensity, and the answer was a louder alarm. The research said otherwise. After a year of building the alternative in Kotlin, my own dose adherence said otherwise too.

This article is the long version of what I learned. The seven-part stack in section three is the part I would most want a developer building a competing product to read, because it is the part the SERP misses.

## Why do most medication reminder apps fail?

The numbers on ADHD medication adherence are worse than most people expect. Anthony Rostain, M.D., a professor of psychiatry at the Perelman School of Medicine at the University of Pennsylvania, has reported that "only 20 to 40 percent of patients follow their medication regimen regularly, if at all, after 12 months of treatment" [Rostain, ADDitude]. He adds in the same piece that "more than two-thirds of patients take their stimulants on only three out of five days, or even less." This is the typical pattern, not the failure case.

A clinical review by Pappadopulos and colleagues, summarized through CHADD, found that for children with stimulant-treated ADHD, "greater than 50% ceasing treatment despite their efficacy" [Pappadopulos et al., via CHADD]. The medication worked. The continuation did not. The failure point is not efficacy. It is the gap between intending to take the dose and actually taking it, every day, indefinitely.

The standard reminder-app answer to this problem is to make the alarm louder, more frequent, or more visually demanding. The research on alert response says this backfires. A 2017 systematic review of clinician alert systems found that response rates dropped roughly 30 percent per repeated alert [PMC 2017 alert habituation review]. A 2022 Frontiers in Digital Health study on SMS medication reminders in adults with ADHD found similar fade-out: alerts lose efficacy within about two weeks of daily use [Frontiers in Digital Health, 2022]. The reminder that worked for the first three days stops working by the third week, and the apps that escalate in response train the dismiss reflex to fire faster.

The deeper point is that medication itself does not fix the part of the brain a reminder is asking to act. Stimulants treat hyperactivity, focus, and impulsivity in chosen contexts. They do not directly improve time management or transition between activities [Hallowell, HealthCentral 2022]. The Cleveland Clinic guide is explicit about this scope [Cleveland Clinic]. The window in which a person needs the reminder to land is the same window in which executive control is least supported. For more on why stimulants do not stop the bingeing loop, see [why your ADHD meds aren't stopping the scroll](https://pause-moment.com/for/adhd/why-adhd-meds-arent-stopping-the-scroll/).

## What does personalization actually change?

If louder is the wrong axis, what is the right one? The answer the research keeps pointing at is salience that does not habituate. A 2020 Vision Research study on visual attention found that novelty and saliency compete for gaze control in priority maps, and that personally relevant stimuli retain salience longer than impersonal ones [Wang & Theeuwes, 2020, Vision Research]. Generic notification design assumes a single repeating cue can keep its grip on attention. The data says no. The data says repeated identical cues fade, and that fade rate is steepest for cues with no personal anchor.

A 2022 randomized study published in Cyberpsychology (N=112) compared design-friction approaches against goal-setting approaches for reducing problematic phone use and found that design friction performed better than intention-based approaches alone [Cyberpsychology 2022 RCT]. The implementation-intentions literature points the same direction: linking a planned action to a specific cue, decided in advance and triggered by an external signal, outperforms intention without external cue [Pirolli et al., 2017, JMIR]. The combined picture is that personal salience plus structural triggering beats generic-alert escalation, by a margin large enough to show up in peer-reviewed effect sizes.

What this means for a medication reminder is that the lever is not the loudness of the alarm. It is the personal anchor of the cue and the structural protection of the moment after the cue fires. Both are dimensions most apps in the category treat as cosmetic upgrades, not as the actual mechanism. Personalization in the apps I tested was always a Settings page option, not the load-bearing element. The thesis of this article is that personalization is the load-bearing element, and the rest of the product should be designed around it.

## The personalization stack

Here is what I mean by a stack. Seven elements, each small on its own, that together describe a different shape of product. For each one I will tell you what stopped working in my own routine and what I built differently. This is the story of what I learned trying every app I could find, not a feature-by-feature comparison.

**1. The photo on the lock screen is your own.** A generic pill icon is the same pill icon for everyone, and my eye stopped registering it within a week. The personal photo does not fade that way. Dosecast has photos, but as small thumbnails next to the medication name; I needed the photo to fill the lock screen, not sit in a list. MyTherapy is a well-designed tracker, but I found the setup pattern did not fit my ADHD: too many fields to fill in before any reminder fired. Pause Moment makes the personal photo the entire lock screen at the moment of the cue. Not a thumbnail. The whole screen.

**2. The words are written by you, not the app.** App-written notification copy is generic by necessity. The version of you that picked the morning and wrote "take it now, one missed day costs three" has context the distracted version of you does not have at medication time. Pause Moment lets you write the words once, when your head is clear, and shows them at the moment your head is not.

**3. The sound is your choice, including silent.** I tried Medisafe for a while. When custom sounds moved behind their paywall, that broke my flow; the alarm I had set was the one that worked for me, and being asked to pay to keep it felt like the wrong gate at the wrong moment. I built Pause Moment with custom sound included from the start, with a default-silent option, because for sensory-sensitive adults silent is the right setting, not a downgrade.

**4. The timer is visible and pre-decided.** Most reminder apps present a notification, and the interaction shape is binary: dismiss it, or do not. There is no defined window in which the moment is held open. Pause Moment shows a visible countdown for the duration you set at setup. You see the timer. You know how long the moment lasts. The structure runs without further input from you, which is the implementation-intention pattern the research keeps validating [Pirolli et al., 2017, JMIR].

**5. The default cue is silent, not alarming.** Pillo's approach did not fit for me. The demanding-alarm framing was the opposite of what my nervous system needed at medication time; the louder the cue, the faster my hand swiped. For some users this is the right product. For me, and for the sensory-sensitive adults the research keeps describing, an alarm-style cue gets filtered as threat. I built Pause Moment with a silent default: the screen state changes, the photo appears, the words appear. The cue is visual and structural.

**6. The lock is genuinely un-dismissable.** This is the part most apps in the category cannot match, because their architecture is built around notifications and notifications are dismissable by definition. Pause Moment ships a scheduled screen lock that holds for the duration you chose at setup. The dismiss-impulse fires, the swipe tries, and the screen stays. The moment is held open long enough for the conscious decision to catch up.

**7. Safe people get through.** A scheduled lock that blocks emergency channels is unusable for parents, caregivers, anyone on-call. Pause Moment lets phone calls ring through during pauses by default. The lock covers the input channels that interrupt without urgency; real-time coordination channels stay open. The same principle carries across the [moments pillar](https://pause-moment.com/for/moments/) for other windows of the day.

## Why the full stack matters together

Individually, each of those seven elements is a small lever. A custom sound on its own would be a Settings page toggle. A visible timer on its own is just UI. A custom photo on its own is a profile picture. The argument of this article is that the seven elements together describe a different shape of product, and the shape is the point.

The pattern is what I think of as a future-you design. The morning version of you, the calm version with clear executive function, sets up the conditions the distracted version will encounter at medication time. You decide the photo. You write the words. You set the duration, the sound, the time. The afternoon version of you, the one whose working memory is shorter and whose dismiss-reflex is faster, runs into the structure the morning-you built. The morning-you has capacity. The afternoon-you gets the benefit.

Most apps in the category assume one configuration fits everyone. The personalization stack assumes the opposite: configuration is the load-bearing piece. A parent of three with ADHD on a morning dose has a different right answer than a graduate student on a wake-time schedule. The shape of the right product is the same. The contents of the lock screen are not. Configuration is the product.

## How Pause Moment fits together

Pause Moment is a scheduled screen lock for Android. You set it up once, when your head is clear, and the structure runs from that point forward.

Open the app. Pick the time your medication is due. Choose the duration: 1, 2, 3, 5, or 10 minutes. Most people pick 1 minute for a daily dose. Pick the photo. Write your words. Choose the sound, or silent. That is the setup, and it lasts.

When the scheduled time arrives, your phone locks. An "I'm Ready" page appears showing the photo and the words you chose. You tap "I'm Ready" to begin the pause; the timer starts. The lock holds for the full duration you chose at setup. The dismiss-impulse fires, the swipe tries, and the screen stays. When the timer ends, you tap "I did it" or "I skipped." The feedback loop is part of the mechanism: the two-button close means the system has honest data about what happened, and you have a record. Notifications keep arriving but stay invisible behind the lock; phone calls come through by default so safe people can reach you.

For ADHD adults whose medication-adherence problem extends into how they spend the focused hours after the dose, see [how to stop scrolling on your work breaks](https://pause-moment.com/for/moments/how-to-stop-scrolling-on-work-breaks/) for the same structural mechanism applied to the post-medication window. The two scopes share an underlying principle: the structure does the work that depleted executive control will not do reliably.

Pause Moment is on [Google Play](https://play.google.com/store/apps/details?id=com.pausemoment.app) for Android. There is a free tier with one reminder. A one-time lifetime tier is also available; see the Play Store for current pricing.

## When this approach isn't enough

Three cases where the personalization stack alone is not the right tool.

**If your ADHD symptoms are severely impacting work, relationships, or safety** despite multiple tools tried, the right next step is an ADHD specialist or your prescribing clinician, not a phone-lock app. Pause Moment addresses one specific failure mode (the dismiss-and-forget loop at medication time). It does not address the broader clinical picture. If you do not yet have a prescriber, that is the conversation to start. For more on what happens when stimulants are doing their job but the bingeing loop runs anyway, see [why your ADHD meds aren't stopping the scroll](https://pause-moment.com/for/adhd/why-adhd-meds-arent-stopping-the-scroll/) and [how to stop ADHD doomscrolling: what the research actually shows works](https://pause-moment.com/for/adhd/how-to-stop-doomscrolling-on-adhd-medication/).

**If co-occurring depression or anxiety is shaping the adherence picture**, a mental-health professional is better positioned than a phone-lock app to address the combined picture. For ADHD adults also on antidepressants who notice the felt sense of presence muted, see [when antidepressants work but you don't feel like yourself: understanding emotional blunting](https://pause-moment.com/for/antidepressants/emotional-blunting-antidepressants/).

**If medication side effects are part of why adherence is hard**, that is a prescribing-clinician conversation. A reminder app cannot help with the side-effect dimension. The medication has to be one you have already decided to take. The lock helps you actually take it.

## Sources

1. Rostain, A. (n.d.). The treatment paradox: better living through chemistry? *ADDitude*. [additudemag.com](https://www.additudemag.com/add-medication-adherence/)
2. Pappadopulos, E., et al. (clinical review summarized via CHADD). Adherence challenges with medications in patients with ADHD. [CHADD](https://chadd.org/adhd-in-the-news/adherence-challenges-with-medications-in-patients-with-adhd/)
3. Frontiers in Digital Health (2022). SMS-based reminder interventions in adults with ADHD and reminder habituation patterns. [Frontiers in Digital Health](https://www.frontiersin.org/journals/digital-health)
4. Ancker, J. S., et al. (2017). Effects of workload, work complexity, and repeated alerts on alert fatigue in a clinical decision support system. *BMC Medical Informatics and Decision Making*. [PMC5470345](https://pmc.ncbi.nlm.nih.gov/articles/PMC5470345/)
5. Cleveland Clinic. ADHD medications: how they work, side effects, and types. [clevelandclinic.org](https://my.clevelandclinic.org/health/treatments/11766-adhd-medication)
6. Hallowell, E. (2022). What ADHD medications do and don't fix. *HealthCentral*. [healthcentral.com](https://www.healthcentral.com/condition/adhd)
7. Lyngs, U., et al. (2022). Supporting people with ADHD in managing problematic smartphone use. *Cyberpsychology, Behavior, and Social Networking*. [Cyberpsychology](https://www.liebertpub.com/journal/cyber)
8. Pirolli, P., et al. (2017). Implementation intention and reminder effects on behavior change in a mobile health system. *Journal of Medical Internet Research*. [JMIR](https://www.jmir.org/2017/11/e397/)
9. Wang, B., & Theeuwes, J. (2020). Novelty competes with saliency for attention. *Vision Research*, 168, 51-60. [ScienceDirect](https://www.sciencedirect.com/science/article/pii/S0042698920300067)

*On AI assistance in drafting this article: This article was drafted with AI assistance, edited and verified by me. The research citations, the lived experience, the product-design decisions, and the personalization-stack thesis are mine. The reader deserves to know which sentences a human stands behind. I stand behind these.*

*This article describes my own experience and the research substrate Pause Moment was built on. It is not medical advice. If you have questions about your specific medication regimen, dose timing, or whether to start, change, or stop a medication, the person to ask is your prescribing clinician or pharmacist.*

If you want the feature-by-feature, app-by-app comparison version of this analysis, see our [best ADHD medication reminder apps for 2026](https://pause-moment.com/best-adhd-medication-reminder-apps-2026/) guide. This piece is the story; the listicle is the comparison.

Ezra Halevi

## FAQ

### Is Pause Moment a medication tracker or a reminder app?

Neither, exactly. It is a phone lock that fires at the time you set, shows your own photo and your own words, and holds the screen for the duration you chose. It does not aim at the complex-tracking job that other apps do well. It does not deliver an escalating alarm. It holds the moment open long enough that the dismiss-impulse loses to the conscious decision to take the dose.

### Why is the lock screen photo such a big deal?

Generic icons habituate. Within a week or two your eye stops registering the pill icon. Your own photo does not fade the same way. The person you started the medication for, or the morning you were proud of remembering, carries personal salience that generic UI cannot replicate. The research on novelty and salience competition is consistent on this point.

### I'm already on Medisafe. Should I switch?

If Medisafe is working for you, do not switch. It is genuinely strong for complex multi-medication regimens and caregiver alerts; that was never the problem I was trying to solve. I built Pause Moment for a narrower failure: the single dismiss-and-forget loop where one missed reminder means one missed dose. If that is your failure mode, the lock-based approach handles it differently. If your failure mode is coordinating eight medications across three family members, Medisafe will serve you better.

### Does the lock screen survive if I close the app?

Yes. The lock is a system-level screen overlay scheduled in advance. Closing the app does not cancel a scheduled pause. The pause fires at the time you set, regardless of what state the app is in. Once you set up the pause, the system runs without further input from you.

### What about emergency calls during a pause?

Phone calls come through during pauses by default. The lock covers the input channels that interrupt without urgency, not the channels that exist for real-time coordination. If a family member, the school, or a healthcare provider needs to reach you during a one-minute medication pause, the call rings through normally.

---
Source: https://pause-moment.com/founder-story/personalization-stack/