I never found efficiency. For five months I imported and categorized my transactions - from payroll through food. I definitely had reports being generated (using hledger) but the amount of time I invested and the ROI wasn’t there.

Had it been 5m a week, maybe I could’ve done it. But it was 3h/wk - finding transactions, transcoding them with scripts I was maintaining, and breaking down the costs with my wife to understand if Costco purchases were “groceries” or “home”.

I nailed the micro and failed the macro. With two kids, and no desire to invest 150h a year into this, there’s no way I can get this to work.

I imagine AI would be great at:

1. Maintaining scripts to extract and normalize various financial artifacts (PDFs, CSVs, XLSXs) into text files

2. Helping draw insights

But manual effort of:

1. Collecting from various banks, paystubs and investment accounts

2. Ensuring flows net to zero and make sense

Is Too Much Work.

▲

rapidfl 3 hours ago | parent | next [-]

The hard part has been to import all the accounts and to categorize the transactions on the 1-2 cards and 1-2 checking accounts that get used a lot. However doing hledger motivated me to simplify things. Either delete cards and accounts. Or barely use them so I can update them once or twice a year. And helps to reduce the number of categories.

I have not made updates for all of 2025 after 3 or 4 years of hledger. ROI diminished once the itch to zero everything wears off and since I have somewhat steady state expenses.

▲

dfee 2 hours ago | parent [-]

If ROI has diminished (maybe dopamine diminished?) then what keeps you motivated?

▲

rapidfl an hour ago | parent [-]

These days, I find it is doable 2-3 times a year. Its a lot of work at one shot but somehow able to catch up so as to review the main spending accounts. The zeroing is what attract me still. I get to see some patterns in my overall spending/categories. And it is side project where there is perceivable concrete progress, so still somewhat easier dopamine.

Simplifying/automating the workflow is a pull. As I type this, I have an idea to stop tracking NW in hledger - it can seem incomplete because money moves from checking to brokerages. That's fine. I only want to see expenses.

	▲	dfee 16 minutes ago \| parent [-]
		Interesting! In a different comment, I took the other side: maybe it’s only worth tracking NW in hledger? Itemizing transactions is heavy duty bookkeeping effort, and I didn’t feel any directional value.

▲

dorgo 3 hours ago | parent | prev | next [-]

have a "Costco purchases" sub-category for "groceries". If you ever feel differently, just move it to "home". Your system has to support this though. I did multiple re-organisations in the last 10+ years.

	▲	dfee 2 hours ago \| parent [-]
		Maybe the right way to do this is to only track accounts, and no tracking of individual txns? Importing and tracking 401k contributions and stock dividends across multiple brokerage accounts - and then trying to handle timing differences (money in flight)… oof

▲

fittingopposite 3 hours ago | parent | prev [-]

Yes. Also wondering about the ROI. What do you gain by this?

	▲	dfee 2 hours ago \| parent [-]
		For transactions: you spot fraud, mischarges or subs to cancel. Really a watchdog service would be good here. High effort, low return. IMO. (And a bit of tracking around reimbursements) For savings: helps you keep an eye on where you’re at by forcing you to reconcile. It’s totally worth it for a low value of invested time. But, it’s not going to be low effort.