Designing LifeMax: Identity-First Habit Systems.
On the difference between tracking what you do and asking who, exactly, you would like to be when the tracking stops.
My mother kept, for many years, a small leather diary in which she recorded — in the looping, faintly disapproving handwriting of a woman who had once trained as a nurse — every glass of water she had drunk in the previous twenty-four hours. She did not, to my knowledge, do anything else with this information. The book lived on top of the microwave, between a pot of basil and the unpaid electricity bill, and once, when I was nine, I asked her what it was for. She looked at me as if I had asked what shoes were for. It's a record, she said, and turned back to the kettle.
I have been thinking about that diary a great deal lately, because I am, against my better judgment, building a habit-tracking application. The application is called LifeMax. It is the third such thing I have built, and I have come, through the slow and humiliating process of using my own products, to a conclusion which I suspect will not endear me to my fellow practitioners in the genre.
It is this. The tracking is not the point. The tracking has never been the point. The tracking is, in fact, the thing that goes wrong.
What the apps are actually selling.
If you open any of the popular habit-tracking apps — and I have, at one time or another, paid money to most of them — you will find, beneath the various badges and streaks and gentle gamified noises, a single, mostly unexamined assumption. The assumption is that what you want, at bottom, is to do a thing more often. To meditate more often. To run more often. To read more often. To drink, in my mother's case, eight glasses of water per day at regular intervals between the hours of seven and nine.
I do not believe this is what people actually want.
What people want, in my experience — what they shyly admit to wanting, when you sit with them long enough that the conversation outlasts the coffee — is to be a slightly different sort of person. Not to do the running. To be a runner. To be the kind of woman who, on a wet Tuesday in November, is not surprised to find herself in trainers. To be, in short, somebody else; or rather, the version of themselves they suspect, in their better moments, has been quietly waiting to be let out of a cupboard.
The apps cannot help with this. The apps can count. Counting and becoming are very different operations.
An evening with my friend Priya.
I had dinner last month with a friend — Priya, who runs a small art gallery on Valencia Street and who has, over the seven years I have known her, started and abandoned approximately eleven attempts at a daily journalling practice. We were eating, as it happens, ramen. She had her phone face-down on the table, which is the universal gesture of a woman who is about to say something she means.
"The thing is," she said, prodding her egg, "I don't really want to journal. I want to be the sort of person who has journalled. There's a difference."
I have stolen this line from her several times since. I will, I imagine, steal it again before this essay is over.
The difference she pointed at — between the doing and the having-done, between the practice and the identity it is supposed to deposit — is the entire crack the existing software falls through. You cannot, by tracking a behaviour, make it feel like an identity. You can only, in fact, make it feel like a chore one happens to be unusually good at logging.
So I am building it backwards.
LifeMax begins, instead, with a question. The question is not what would you like to do this week?. The question is who would you like to be on a Sunday in three years' time? — and the answer is supposed to be a sentence, not a number. The sentence is supposed to be specific enough to embarrass you slightly. Mine, at the moment, is: a man who reads in the mornings and is good to his sister. Priya's, last I heard, was: a woman whose flat does not look as if it has been burgled by an enthusiastic cat.
From the sentence, the application — politely, and with what I hope is a certain restraint — proposes a small number of practices that the sort of person in the sentence might already be doing. It does not, crucially, give you a streak. Streaks are how you get a man who has read for thirty-one consecutive mornings and is no kinder to his sister than he was on day one. Streaks measure the wrong thing, beautifully.
What the application gives you instead is a weekly question, on Sunday evening, and the question is always the same. Did this week feel like the kind of week the person in the sentence would have had? You answer yes or no. Over time, the application learns the shape of your particular yeses, and asks, gently, when the noes begin to stack up, whether perhaps the sentence has changed. It very often has. People are, in my limited experience, mostly trying to become someone they have already half-outgrown.
Why I think this matters.
I do not, on the whole, believe that software can make you a better person. Software can, at best, get out of the way of the better person you were already trying to be. The existing apps do not get out of the way. The existing apps stand directly in the way, holding a clipboard, pleasantly demanding that you log your water.
If LifeMax works — if it works at all, which is by no means certain, since most of what I build at three in the morning does not — it will work because it begins from the right question, and because it has the manners to stop talking once the question is asked.
My mother, were she still alive, would not, I think, have used it. She had her diary, and she had her kettle, and she had — at the level beneath the diary, beneath even the kettle — a perfectly clear sense of who she was trying to be: a woman who had, against considerable odds, raised a son who would one day write essays about her. On that, the record will show, she did not need an app at all.
— B.