← Blog

Sparky vs one sec: A Pause Before Opening, or Minutes Earned

October 30, 2026

I build Sparky, so weigh this with that in mind. The short answer: one sec has published evidence and Sparky has none. one sec puts a short pause and a way out in front of every tap on an app you chose. Sparky keeps the app shut until you spend minutes you earned from habits. Pick one sec if a pause is enough for you. Pick Sparky if you want the app to cost something.

Below, I read one sec’s two studies against each other and against its App Store listing. Whether earning screen time works at all is a separate question, covered in our evidence write-up.

How do one sec and Sparky stop you from opening an app?

one sec asks every time. Each tap on a chosen app plays a breathing animation, 6 seconds by default, then offers two choices: go on, or go back. Sparky asks once per session. Its shield offers “Start Time Bank” or “Stay Paused”, and a started session opens your gated apps for up to 60 minutes.

In one sec, the wait can run from 3 to 60 seconds, according to the CHI 2024 paper. The one sec listing also offers a “Re-Intervention”: “After a set time (e.g. 5 minutes) you’ll be kicked out of social media, and you’ll have to go through an intervention again.” Using one sec with one app is free.

Sparky, the app where you earn screen time by finishing habits, prices minutes in Energy. In Sparky 1.6, the App Store version on 9 October 2026, one Energy buys one minute. Default habits pay 1 to 10 Energy per check-in, and the daily earn cap is the one you pick at setup: 60, 120 or 240 (HabitEconomyModels.swift, OnboardingViewModel.swift). A Sparky session runs by the clock, at most 60 minutes, and stopping early returns the unused minutes (AccessBankService.swift). Inside a session, reopening Instagram meets no question. Outside one, every tap meets the shield.

What did the 2023 PNAS study of one sec find?

The PNAS paper by Grüning, Riedel and Lorenz-Spreen followed new one sec users for six weeks. People closed the target app again on 36% of their attempts. The paper also reports that “one sec decreased users’ actual opening of target apps by 57% after six consecutive weeks.” Co-author Frederik Riedel built one sec, and the paper’s competing-interest note says so.

The weekly numbers show the shape. In week one, people dismissed 43% of their attempts. By week three it was 33%, and it stayed between 32% and 34% to the end. Attempts fell from 166 per user in week one to 105 in week six. Participants also reported spending 77 minutes a day less on their apps, a self-reported figure, not a measured one.

Two details are easy to miss. The paper recruited 719 new users, and only the 280 who used one sec for all six weeks are in the main analysis. The authors say a version with dropouts, in their supplement, gives similar results. And the 57% counts app openings, not minutes. By my arithmetic, 105 attempts with about a third dismissed leaves about 70 openings, against 166 attempts in week one. That is the 57%. The listing says “app usage drops by 57% on average!” The paper’s number is real, but it measures openings among the people who stayed six weeks.

What did the 2024 CHI study add?

The CHI paper by Haliburton, Grüning, Riedel, Schmidt and Terzimehić studied 1,039 people who already used one sec, over an average of 13.4 weeks. It logged 1,318,331 interrupted app opens. People went on to the app 67.0% of the time and dismissed it 33.0% of the time. Both attempts and dismissals fell over the months.

The authors read the falling dismissal rate as a good sign: fewer reaches, and the ones left are on purpose. One participant (P162) described the other reading: “my brain has gotten used to the waiting before an app opens. It is no longer an inconvenience.” The attempt curve, the authors note, flattens after about 20 days.

The CHI paper also found breaks. 280 users (26.9%) had at least one gap of more than 48.2 hours with no one sec pause logged, which the authors count as a break. The 970 breaks averaged 9.15 days, and most often started on a Friday and ended on a Sunday. The authors conclude that “the impact of the design friction does not last beyond its application.”

The CHI study has no control group, and its data came from people still using one sec, so it “is inherently not possible to capture dropout patterns,” in the authors’ words. Riedel is again a co-author, with riedel.wtf GmbH as his affiliation. The App Store lists one sec’s seller as riedel.wtf apps S.L.

What do the two papers show when you read them together?

Together, the papers say the pause works while it is on, stops a smaller share of taps as weeks pass, and was measured on the people who stayed. That is still more than any earn app has. It is not proof that one sec cuts screen time for everyone who installs it.

Neither study counts the people who left. PNAS set aside 439 of its 719 new users (61%) from the main analysis, and CHI could not see dropouts by design. Both show the share of taps stopped falling, then settling. And both tie the effect to the pause being on: PNAS measured people who kept it on, and CHI watched what happened when people turned it off.

The app has also changed since. The PNAS paper describes a 10-second delay, CHI found most users on 6 seconds, and on iOS 27 one sec can now import its Shortcuts automation in one step (on iOS 26 the setup is still by hand, per one sec’s own guides). The studies test a pause with a way out, not the exact build you would install today.

A break from the pause is not a break from the apps. In Lambert et al. 2022, 154 adults were randomly assigned to a week off social media or to their usual use, and the break group’s well-being improved. CHI’s breaks were the opposite: the apps stayed, and the brake came off.

How does Sparky’s gate compare with one sec’s pause?

Sparky has the same way out but offers it less often. Every Sparky shield has a “Stay Paused” button that closes the app, and the PNAS online experiment found that the option to say no did the most work. Sparky shows it only while no session runs. A session removes the question for up to an hour.

Put a number on it. In the PNAS study, a new user tried to open a target app 166 times in week one, about 24 a day. With one sec, each of those taps meets a pause. With Sparky, taps inside a session meet nothing. One 30-minute evening session could absorb several of those reaches without a question.

That is where one sec is the better tool. If your problem is many quick checks across the day, a pause on each one fits it. Sparky bets on something else: minutes that come from habits, a clock that runs whether or not you scroll, and unused minutes back in the bank. Nobody has measured whether a price wears in the way the wait did for CHI’s participant.

We cannot measure it yet either. Sparky keeps habits and shield taps on the phone, so we cannot count how often people tap “Stay Paused”. The tracking Sparky’s privacy label declares comes from install attribution (Atlas), which starts after Apple’s App Tracking Transparency prompt. We cover what each label means in what screen time apps can see.

Has anyone tested earning screen time the way one sec was tested?

Not for a phone app. As of 9 October 2026, I found no peer-reviewed randomized trial of an earn-your-screen-time phone app, Sparky included. Our evidence write-up has the searches and the preprints. The nearest peer-reviewed result earned TV and tablet time, not phone time.

That trial is StandUPTV (Keadle et al., 2025). 110 adults were randomized across eight combinations of three add-ons for 16 weeks. In the earn add-on, 10 minutes of Fitbit-measured exercise bought 30 minutes of TV or tablet time. The version with the core app plus earning cut recreational screen time by an expected 118.1 minutes a day (95% CI 73.1 to 163.0). The core app alone cut 88.2 minutes a day, and the two ranges overlap widely, so not all of the 118.1 minutes belongs to earning. The authors report that some participants found the ratio too lenient. The groups are small, and StandUPTV tested a 1-to-3 exercise rate on a TV and a tablet, not habits on a phone. It is evidence about earning, not about Sparky.

What do one sec and Sparky cost, and what do they collect?

On 9 October 2026, one sec was free for one app. one sec pro cost $6.99 a month, $19.99 a year or $99.99 for life, with family plans at $9.99 a month, $29.99 a year or $149.00 for life. Sparky is free with five active habits. Sparky Pro costs $8.99 a month, $59.99 a year or $129.99 for life.

one sec’s listing showed version 6.0.2 (29 September 2026), 4.83 stars from 23,669 ratings, a 385.3 MB download and iOS 15.0 or later. Sparky is 59.9 MB, needs iOS 26.0 and has no ratings yet.

The privacy labels differ more than the prices. one sec’s label declares no tracking and lists one item, Usage Data for analytics, as not linked to you. Sparky’s label declares Purchases, Identifiers and Usage Data as used to track you, for the attribution reason above.

Which one should you pick?

Pick by the shape of your habit. one sec suits many short reaches across the day, people who want published evidence, and anyone whose iPhone cannot run iOS 26. Sparky suits people who want screen time tied to habits, minutes that carry over, and a closed app with no countdown to sit through.

The one sec studies have limits, and they are still the best evidence any app in this category has. If you want the habit-priced version instead, Sparky is free to start, and more earn apps are compared in our guide to apps that make you earn screen time.

Sparky turns small habits into screen time you earn. Download it free on the App Store.