The hardest part of writing software is always the requirements gathering.
I came to Tiller with a “Future Checkbook” spreadsheet that I was able to quickly tiller-ize the inputs on and reduce my manual work, but I thought “why can’t these entries be automatic?” And started writing a version that automatically detects my recurring payments. Sure, on the surface of it all the math works, but reality isn’t always as neat.
- New bills or annual bills without a 2 year history don’t have enough information.
- Have to add a section for manual additions
- Bills with the same name but 2 occurrences with different schedules can’t figure out the schedule from averages.
- I guess the manual section needs to override the automatic section.
- Maybe I bought paint 5 times during a renovation, that doesn’t make it a regularly reoccurring bill, and sometimes actual recurring bills stop/change.
- Filtering those by # missed payments
- Still need manual entry of manual payments not yet cleared
- There’s no universal rule for which items should be budgeted at the max, latest, or average amounts.
There’s so many exceptions to the rules, I’m thinking of just sticking with my version with manually listed recurring payments that’s been 23 years in development and just has a little help now from the live data. There’s already a few projection sheets out there, so no need to share mine, it’s just what I’m used to, not better.
I did learn a lot though about google sheets, so it was not a waste of time to try to build it.
Progress on automatic report:
. Well, if you have a breakthrough, others doing something similar will certainly want to see it!