I used Mint for about three years before I gave it up, and for most of that time I would have told you, sincerely, that it was working. My accounts were linked. The categorisation engine was reasonably accurate after I had trained it. I could open the app on a Tuesday morning and see, in colour-coded segments, exactly where my money had gone the previous month. By every visible measure, I was budgeting. When Intuit announced in late 2023 that Mint would be shut down — formally retired by March 2024 — I felt the small panic that the announcement seemed designed to produce, and I spent an evening looking at the available alternatives. Then I closed the laptop, opened a notebook, and wrote down the four kakeibo questions instead. That was the moment I realised what Mint had actually been doing to my financial attention, which was: removing it.
This essay is about why kakeibo beat Mint for me, and why I think the same will be true for many of the readers who came to budgeting through Mint and were left, abruptly, looking for somewhere else to land. The thesis is short. Automation removed the attention that makes budgeting work. The friction was the feature.
What Mint actually did
Mint did three things well, and it did one thing badly, and the thing it did badly turned out to be the thing that mattered most.
The three things it did well were aggregation, categorisation, and visualisation. It pulled transactions from every account I owned, sorted them into categories that were mostly correct, and displayed them in tidy charts I could consult at any moment. These are real capabilities and I do not want to dismiss them. For a household that has never in its life sat down with its own numbers, Mint provided an instant, painless overview that no paper ledger could match.
The thing it did badly was that it did all of this without me. I was a spectator to my own financial life. The transactions appeared. The categories were assigned. The charts updated. My role was to look at the result and to feel either pleased or vaguely uneasy, and the difference between those two reactions almost never produced a behaviour change. I would see, on a Tuesday, that the dining-out category had run high in October. I would feel a flicker of something. By Wednesday morning the flicker had passed and I was back to behaviours that produced an even higher dining-out category in November. Mint had told me. I had not actually heard it.
This is not a flaw of Mint specifically. It is a structural property of any auto-budgeting tool. When the act of recording requires no effort, the act of recording teaches you nothing. You cannot learn from a record you did not write. The automation is not neutral; it is anti-attention.
The 2024 shutdown and what it revealed
When Intuit announced Mint's closure, the reaction in personal-finance forums was instructive. Some users were angry. Some were resigned. A surprising number, when they actually looked at their accounts after the announcement, discovered they had no clear sense of what their finances had been doing for the previous year. They had a tool that displayed everything and a memory of almost nothing.
I include myself in that surprising number. When I went looking for what Mint had taught me about my spending across three years of use, I found that I could recite vague trends — I spend more on dining out than I want to — but I could not produce a single specific behaviour change I had made on the basis of what the app had shown me. Three years. No specific change. The app had shown me my money continuously. I had absorbed almost none of it.
This was not Mint's fault. It was a misunderstanding I had about what budgeting is. I had thought budgeting was a problem of information — if I could just see my spending clearly, I would change it. Mint solved the information problem completely and the spending did not change. What I had actually needed, and what kakeibo provides, is not information. It is a structured weekly and monthly act of looking, by hand, at numbers I have written myself.
What changed when I switched
I will be specific about what was different in the first three months of kakeibo, because the abstract claim — handwriting produces attention — only lands if it is anchored to actual behaviour.
In the first month, what changed was small. I logged everything by hand for thirty days. The act took, on average, about three minutes per evening. By the end of the month I had a notebook full of small entries that, taken together, told me three things I had not known despite three years of Mint.
The first was that my optional category was bigger than I had thought, and the size came not from a few large purchases but from many small ones. Mint had shown me the total but not the texture. The notebook showed me the texture by forcing me to write each entry, and the texture was where the real story lived.
The second was that I had two specific spending triggers — Friday evenings after work, and Sunday afternoons when I felt restless — that were producing roughly 40 percent of my optional spending. Mint had aggregated by category. The notebook was implicitly chronological. The triggers were obvious within two weeks of writing the entries by hand.
The third was that my actual savings rate, calculated honestly from intake to month-end, was lower than the figure Mint had been showing me. Mint had counted certain transfers as savings that were not really savings — money that had moved into an account I would draw from again within ninety days. The notebook, with its blunter categories, told me the truer number.
By month three, two of the three patterns had changed materially. Not because I had set goals or used willpower, but because I had seen them, in my own handwriting, in a way I could not unsee. This is the mechanism. It is not magic. It is the small implementation pause that handwriting produces, accumulated across thirty days, against a record I have to look at.
The handwriting evidence
I want to mention briefly the research, because some readers will find it useful and others will already trust the experiential claim. A 2014 study by Pam Mueller and Daniel Oppenheimer in Psychological Science compared note-taking by hand and by laptop and found that handwriting produced significantly deeper conceptual processing of the material, even when the typed notes were more complete. Subsequent work has extended the finding to other domains. Handwriting forces selection; selection forces engagement; engagement produces memory. Mint's automation removed all three steps. Kakeibo restores them.
This is also why digital kakeibo tools — including the Notion ledger I now use for the running tally — only work when paired with manual entry. I type each transaction myself. I do not import from my bank. The five seconds of typing is not a tax on the practice; it is the practice. A digital ledger that auto-imports is just Mint with a different interface.
What replaces Mint, properly
If you came to this essay because Mint shut down and you are looking for what to do next, I would offer the following sequence, which is what I tell readers who write in.
Begin with a paper notebook for the first month. Not because paper is sacred, but because the resistance you will feel is diagnostic. If you cannot maintain a thirty-day paper log, the problem is not the tool you are choosing; it is that you have not yet built the daily attention budgeting requires. A digital tool will let you fail invisibly. Paper will not.
After the first month, if the practice has taken hold, decide whether you prefer paper or digital for the long term. Both work. I keep both — paper for the monthly reflection, Notion for the running tally — and the dual format suits me. Many readers prefer one or the other. The choice does not matter as much as the rule that you enter every transaction yourself.
Do not, in the first six months, use any tool that connects to your bank accounts. The connection is what allowed Mint to hollow out my attention. It is a feature that looks like convenience and functions like sedation. After six months, when the practice is established and you can feel the behavioural muscle, you may decide a partial bank connection is useful for fixed costs. Even then, log all variable spending by hand.
For readers who want a structured starting point, the longer essay on the Japanese budget method walks through the system end to end, and the companion essay on the four questions covers the monthly reflection in detail. The comparison piece on kakeibo versus YNAB may also be useful for anyone who is considering YNAB as their post-Mint replacement, since YNAB is, in my view, the strongest of the auto-import tools but still suffers from the structural problem this essay describes.
One last thing
I do not want to overstate the case against automation. Mint helped a lot of households start paying any attention at all to money, and that is not nothing. The problem is that paying attention to a chart someone else made for you, drawn from data you did not enter, is a thinner kind of attention than the kind that produces lasting behaviour change. Three years of Mint produced no specific change in my spending. Three months of kakeibo produced two. The difference is not effort — kakeibo is, in clock time, easier than I had expected. The difference is that kakeibo asks you to be present, briefly, every evening, with the small concrete fact of what you spent today. That presence is the practice. Mint removed it. Kakeibo restored it. When Mint shut down, I was, in the end, grateful — it forced me to stop confusing the appearance of attention for the thing itself. If you are in the same place, the notebook is on the desk waiting. Begin tonight. The system you actually need will arrive within a month, and it will not require a subscription to keep working.
If you would like the occasional Tokyo dispatch about the practice as I keep it, I write a short note most months on the Mindful Yen Substack.