Ask someone who is genuinely on top of their money how they do it, and a surprising number will admit it comes down to a spreadsheet. Not an app with a waiting list. A grid, maintained by hand, usually a bit ugly.
This is easy to read as a failure of tooling, and mostly it is. But it is worth asking why the grid keeps winning. Each finance app tends to understand one part of your money extremely well — the current account, the card, the portfolio, the bills. The spreadsheet understands none of them well, and all of them together. When the question is "where do I actually stand", the second thing turns out to be more useful than the first.
The grid has a second quality that is harder to name: it does not argue with you. It never re-categorises a transaction overnight, never quietly restates last year in this year's terms, never decides a payment was a transfer. Whatever it shows you, you put there. For something as personal as money, that predictability is worth more than a lot of automation.
Trove is an attempt to keep that sense of ownership and then add the things a grid genuinely cannot do: a little structure, a little memory, and a clearer view of what is changing over time. Categorisation should be a suggestion you can overrule, not a verdict handed down. History should stay true — a transaction from eight months ago keeps the exchange rate it actually happened at, because restating the past with today's rate is a quiet form of lying.
None of this is a finished argument. It is the belief the product is being built on, and the part most worth getting wrong in public.