How it started

Alex had been trying to start a journaling practice since his late twenties. The attempts followed a pattern recognizable enough that he could name it before the next one failed: he would buy a new notebook or open a new app, write two or three entries that felt thoughtful and genuine, and then skip a day because he had nothing interesting to say. The skip turned into a week. The notebook moved to a drawer. He blamed perfectionism when he talked about it, and perfectionism was accurate as far as it went, but the more specific version was that he had absorbed, from years of writing for clients, a professional standard for what written words were supposed to do. They were supposed to arrive at something. They were supposed to have a point. A journal entry that said he was tired and irritated and had no particular insight into why felt, in his hands, like a draft that had not earned the right to exist yet. He had tried free-writing prompts, structured formats, a five-year diary that asked for only three sentences per day. He had tried the morning-pages method, which lasted eleven days before he decided the entries were garbage and stopped generating them. He was aware that calling his own journal entries garbage was a problem, that the practice was not supposed to produce polished prose, that the whole point was accumulation over performance. He knew this. He continued to delete entries anyway. By the seventh attempt, in his early thirties, he had produced four entries over nine days, deleted two of them, and abandoned the notebook in a coffee shop on Valencia Street, deliberately, without going back for it. It was a fourteen-dollar leather-bound notebook. He left it there.

The contract

$3/day staked against daily journaling, charity: mental health.

In early spring, Alex set up a Lockin contract built around a learning challenge. He whitelisted Day One — the journaling app he had bought, abandoned, redownloaded, and abandoned again twice in the previous three years. The contract's terms were simple: the operating system had to register at least 10 minutes of active time in Day One on his phone, accumulated in any combination, before the daily deadline he had set during contract creation. Active time meant the app open and in the foreground; background minutes and locked-screen time did not count. The deadline was a step in the wizard he had spent more time on than he expected. The time picker had defaulted to 11:59pm, and he had pulled it back to 9:00pm before he confirmed the contract. The tighter deadline raised the daily reward, which he liked the framing of, but the larger reason was that he knew himself: anything past ten started to feel like a free pass, and a free pass was how the previous seven attempts had ended. Once the contract was confirmed, the 9:00pm deadline applied to every scheduled day for its duration — there was no edit-mid-contract option, which was part of why the choice had taken him a while. He chose a mental-health charity as the forfeit destination. The amount was $3 per day — small enough that it would not feel punitive on a hard week, large enough that it had a name in the weekly budget. He picked the learning-challenge structure deliberately. The phone was where his attention already lived. If he could route ten minutes of it into Day One every evening, the practice would happen — Screen Time did not check what he wrote, but the app open in his hand, the cursor blinking on a fresh entry, the absence of a TikTok feed all functioned as a single environmental commitment device. The first eight days went without incident. Day one he wrote 200 words about a project he was anxious about, watched the daily app-time chip on his Lockin card creep from 0:00 to 12:18 over the course of a single sitting, then closed the app and went to bed. Day two was shorter — a paragraph about a conversation with a friend, three sentences about what he had eaten, a note about the weather that he almost deleted and then left in. Day three he wrote for eight minutes about nothing in particular, feeling slightly ridiculous, which was different from feeling like the entry was not good enough to keep. Days four through eight followed the same pattern: not insightful, not polished, but present. The Lockin chip cleared its 10-minute mark on every one of those days, often by accident, while he was still typing. He was keeping a record. He had not kept one in six years. The contract held.

The night it almost broke

Day 9 was a Tuesday. The day had been hard but unspecific, which was a particular kind of hard that resisted the kind of writing he could let himself keep. A client had sent three rounds of revision notes on copy he thought was finished, the notes contradicting each other in ways he was too tired to diplomatically flag. Lunch with his sister had surfaced an old tension neither of them named directly. He had gone to bed at 11pm three nights running and woken up still tired. By 7:15pm he was on the couch with his phone face-down on the coffee table, telling himself he would open Day One in fifteen minutes. The Lockin card showed 105 minutes until the 9pm cutoff and the daily app-time chip stuck at 0:00. At 7:30pm he was thinking about what he would write. He drafted a sentence in his head — today was a hard day in a way that — and rejected it. He drafted another about the client and rejected that. He drafted two sentences about his sister, which felt accurate, then felt invasive, then dissolved before he could put them on a page. He drafted a sentence about the insomnia. The sentence sounded like self-pity, which was a category his internal writing editor treated as unacceptable, and he discarded it. At 8:00pm he was still on the couch. He had now drafted and rejected sentences for forty-five minutes. The phone was an arm's length away, the Day One icon on his second home screen, the daily chip on his Lockin card still reading 0:00 with 60 minutes left until the 9pm deadline. He was aware that what he was doing was the thing he had always done, the thing that had killed seven previous attempts. He was waiting for a sentence good enough to justify opening the app, and opening the app did not require a good sentence — it required a thumb on the icon. The two parts of his mind were in different rooms. At 8:38pm he picked up the phone. He had Day One open for fourteen seconds — long enough to see the prompt screen — before he closed it. The chip on his Lockin card ticked from 0:00 to 0:14 and then froze when he switched back to the home screen. He could not produce ten minutes of active time inside the app between 8:38pm and 9:00pm without typing into it, and typing required a sentence he had not yet allowed himself to release. The arithmetic had collapsed somewhere between 8:32pm and 8:38pm and he had watched it collapse without typing a single word. By 9:00pm the chip read 0:14. The contract closed at midnight. The forfeit registered: $3 to mental health.

What it cost

Alex opened the Lockin dashboard the next morning with his first coffee. The forfeit line was there, date-stamped, amount confirmed: 14 seconds of Day One time before the 9pm cutoff, well short of the 10-minute minimum. He looked at it for a while. He was not surprised. He was not particularly angry at himself. What he felt, sitting with the number, was a specific clarity about what had happened the night before that was harder to access in the moment at 8:38pm when he was setting the phone back down on the coffee table. The entry did not need to be good. That had been the premise of the contract. Ten minutes of active time inside Day One, no quality threshold specified anywhere in the terms. The OS counter did not check what was on the screen. It only checked whether Day One had been the foreground app for at least 600 seconds, accumulated in any combination, before 9pm. He had imported a quality threshold from somewhere outside the contract, applied it to every sentence he tried to draft in his head before opening the app, and rejected everything that failed the test. The test was not part of the agreement he had made with himself. He had added it in real time, between 7:15pm and 8:38pm, when he was tired and the day had been hard and the empty entry screen was easier to leave closed than to fill with something that felt inadequate. The $3 was not large. What was legible in the number was the pattern: this was the seventh attempt, and the seventh attempt had run for nine days before the same mechanism ended it. Not external circumstance. Not a missing notebook. The mechanism was internal, repeatable, and had now cost him something measurable.

Forfeit

$3 → mental health

What changed

Alex reset the contract the next morning with a single rule change, layered on top of the learning challenge. Once Day One was open, the first three sentences he typed had to be honest. No quality threshold. The word honest was doing specific work: it ruled out performed insight and it ruled out backspacing. If a sentence was true when he wrote it, it stayed in the entry. He did not have to like it. He did not have to read it back. He had to type it and leave it there and let the OS counter accumulate the rest of the ten minutes around it. Day 10: he opened Day One at 7:02pm with no sentence in his head and typed three sentences about feeling tired and missing the version of his sister who used to live in the apartment on Deakin Street before she moved across the bay. The Lockin card chip began counting up from 0:00. He kept the app open and added a paragraph about a song that had been in his head, then a list of the week's projects, watching the chip climb to 11:42 before he closed it. Day 11: a list of the week's active projects and which ones he was avoiding. Day 12: a paragraph about a Chet Baker song that had been in his head since Monday, which he described without explaining why it was there, because he did not know why and the contract did not require him to find out. The entries got longer on their own over the following weeks. Not because he was trying to write more, but because once Day One was open and three honest sentences were on the screen, three honest sentences often produced a fourth. Six months in, he had 178 entries. He read through them on a Sunday afternoon and found that the entries he would have described as insightful were not the ones he had planned to be insightful. They were the ones he had written planning to produce garbage and had let himself write anyway. The couch version of his practice had required him to arrive at a sentence before he was allowed to open the app. The app-time version required him only to open the app. The difference was the only thing that had ever made the practice work.

Three sentences drafted. Three rejected. He never opened the app.

How it started

The contract

The night it almost broke

What it cost

What changed

Try the same contract.

Other forfeit stories

Stop deciding. Start staking.