Three years ago I planned a fortnight abroad and optimised every individual booking for the lowest available price. Cheapest flights, cheapest accommodation, cheapest ground transport.
When I added everything up afterwards, including the things I had not counted, the trip cost more than the version I had rejected as too expensive.
Here is where the money went, because the pattern generalises.
The airport that was not the airport
The flight saved a meaningful amount by using a secondary airport.
The transfer from that airport to the city was ninety minutes and cost about a third of the flight saving, each way. It also ran infrequently, which meant a long wait on arrival and an uncomfortably early departure at the end.
The early departure meant an extra night's accommodation near the airport, which I had not counted at all.
Net saving after all of that: close to nothing, plus five hours of my holiday spent on buses.
The accommodation on the edge
Considerably cheaper than anything central, which is why I booked it.
What it cost was two transport journeys a day for eleven days. Individually trivial, collectively about a third of the accommodation saving.
More expensively, it changed the shape of the days. Going back to drop things off was impractical, so I carried everything. Coming back for a rest in the afternoon was impractical, so I did not.
And in the evenings, the cost of getting back late meant I left places earlier than I wanted to, which is the part I actually regret.
Eating at the wrong times
The accommodation had no kitchen, which is fine, and I had assumed I would eat cheaply.
What happened is that I was out all day, got hungry at inconvenient moments in inconvenient places, and ate whatever was nearby. Which is the most expensive way to eat anywhere, because you are buying convenience under time pressure.
The pattern was consistent: a cheap breakfast, an expensive and unsatisfying lunch bought in a hurry near something I was looking at, and an evening meal chosen because it was on the route back rather than because I wanted it.
Somebody staying centrally goes back, eats at their own pace, and spends less.
The tickets I bought twice
A specific and avoidable error. I bought individual transport tickets each time rather than a pass, because each individual ticket was cheap.
Over eleven days the total was roughly double what a weekly pass would have cost. I knew passes existed and did not do the arithmetic because each individual purchase felt too small to think about.
This is a general failure mode with cheap trips. The savings are visible in the large decisions and the losses accumulate in decisions too small to notice.
What I did not do
The real cost, and the one that does not appear in any total.
There were two things I had wanted to do that I did not, because by the time I got round to them I had spent more than I intended and was economising.
Both were the sort of thing I had gone there for. Both cost less than what I had wasted on transfers.
Which is the actual failure. I optimised the cost of being there and underfunded the reason for being there.
What I do differently now
Cost the whole thing, including transfers, transport, and the meals that the arrangement forces on you. On a spreadsheet, roughly, before booking anything.
Treat location as a purchase rather than as a preference. Central accommodation is not an indulgence, it is a transport and time purchase with a measurable value.
Buy the pass on day one, always, in any city where I will use transport more than twice a day.
And set aside a fixed amount for the things I came for, before optimising anything else. That budget is not available for savings.
Where cheap actually works
To be clear, I still travel cheaply and I do not think the principle is wrong.
What works: travelling in shoulder season, which reduces every cost simultaneously with no compromise in convenience. Staying longer in fewer places, which reduces transport and lets you find where local people eat. Self-catering with a real kitchen. And going somewhere that is inexpensive rather than economising in somewhere that is not.
What does not work is taking an expensive trip and grinding down each component, because the grinding costs time and convenience, and time and convenience turn back into money in ways you do not see until afterwards.
The comparison I should have run
What would have settled it before departure is a single figure: total cost divided by usable hours in the place.
The expensive version had a direct flight, central accommodation, and fewer moving parts. It cost more and delivered roughly two more usable days out of fourteen.
On that measure it was cheaper, and the calculation takes about ten minutes on the back of an envelope.
The version of this that is not about money
Some of what I gave up was not financial. Choosing the cheapest option at every point means a lot of small decisions made against yourself.
Walking further than you want to. Waiting for a cheaper connection. Turning down something because of what it costs.
None of those is significant alone and by day nine they had accumulated into a slightly grim mood that I remember more clearly than anything I saw.