What my share of Japan actually cost, and why my own bank data can't tell me
Four of us went to Japan for sixteen days in June — Tokyo, Kanazawa, Osaka, Kyoto, in that order. I went back through my own transaction export afterwards expecting a clean answer to "what did that cost me." What I got instead was a good demonstration of exactly how a personal finance app breaks down the moment a cost stops being personal.
The number that shows up
Filtering to the actual dates I was in the country: £2,732 across 147 transactions. Sounds like a real answer. It isn't one.
Break it down and it's almost entirely small stuff. Something like thirty convenience-store stops — FamilyMart, 7-Eleven, and one recurring entry my own bank insists on labelling "Lawson Animal Feeds," which is almost certainly just Lawson, the actual Japanese convenience store chain, matched to the nearest-sounding business in a UK merchant database and never corrected. Ten Uber rides, averaging around £14 each. And a baker's dozen of larger, entirely unlabelled purchases — no merchant name attached to any of them — that between them account for more than a thousand pounds of the total. A year on, I genuinely can't tell you what half of those were.
What's conspicuously not in there
No flights. No hotel charges, across four cities and sixteen nights. No Shinkansen bookings, no theme park tickets. Every one of those categories was priced in the group's own planning notes before we left — accommodation and transport alone were budgeted at somewhere around £960 a head for the trip — and none of it touched this account directly.
That's not a mistake in the data. It's what actually happened. Four people on one trip means everyone's card gets used for something, and the group squares up afterwards rather than everyone paying their own quarter of every bill in real time. My share of the big-ticket items came out of other people's cards first and got settled between us with transfers over the following weeks — the same weeks in which something like three hundred other transfers were also moving in and out of my account for entirely unrelated reasons: rent, savings, ordinary life. I have no reliable way, looking back now, of pointing at any one of those transfers and saying "that one was Kyoto."
Why this is the opposite lesson from the "full year" post
I wrote up a full year of my own spending a little while ago and the finding was reassuring: tracked honestly, solo spending is a pretty faithful record of where the money actually goes. A single sixteen-day trip with three other people undid more of that record than the rest of the year combined — not because anyone did anything wrong, but because "my spending" and "my share of a shared cost" are different things, and no amount of automatic categorisation turns one into the other after the fact. An app like this is built on the assumption that your outflows are your consumption. Split a cost four ways and pay each other back informally, and that assumption just stops holding.
What the honest number actually is
Somewhere north of £2,732, once my actual share of the flights and the fortnight of hotels gets added on top — probably meaningfully more, given the group's own accommodation-and-transport estimate alone. I can't get more precise than that, and the reason I can't is the finding, not a gap in it. If I want a real total for a trip like this next time, it won't come from reconciling a bank export afterwards. It has to be tracked at the time, by the group, on purpose — which we didn't do beyond paying each other back as it came up.
Not financial advice — I'm not qualified to give any. Figures above are drawn from my own transaction export for the trip dates; I've deliberately left out who I travelled with, where we stayed, and anything from the planning itself beyond the headline budget.