Planning removes uncertainty and it also removes the ability to change your mind. A trip booked to the hour cannot absorb a wet morning, a delayed train, unexpected tiredness or a place you want to stay in longer, all of which happen.
How much should be fixed?
Roughly two thirds, and the two thirds that genuinely need it: accommodation, transport between places, and anything with capped numbers. The remaining third is the days themselves, which are better decided in the morning.
What must be booked?
Accommodation in high season, since availability rather than price is the constraint. Long-distance transport, for price and for compulsory reservations. Sites with genuinely capped entry. That list is shorter than most people assume.
What should be left?
Which museum on which day, where to eat except for one or two evenings that matter, day trips, and anything weather dependent. All of these are better decided the night before with a forecast in front of you.
What is the cost of over-planning?
Obligation. Four timed bookings in a day turns a holiday into a schedule, and the feeling of being behind your own itinerary is a poor substitute for enjoying where you are. Cancellation fees add to it.
What is the cost of under-planning?
Real too, and less common. Arriving in a popular city in August with nowhere to stay, missing a site because it was sold out, or wasting a day deciding what to do are all genuine failures of planning rather than charming spontaneity.
What is a buffer day?
A day with nothing in it, placed before a flight home or after a long travel day. It absorbs delays, cancelled ferries and tiredness, and on any trip involving islands or remote areas it is close to essential.
How do you plan flexibly?
Make a list rather than a schedule. Write what you would like to do in each place without assigning days, then choose each morning based on the weather and how you feel. The list means nothing is forgotten and the absence of dates means nothing is forced.
Does it depend on the destination?
Considerably. Popular cities in summer need more booking; rural regions and shoulder seasons need almost none. A week in Tuscany in October can be planned on arrival; a week in Venice in August cannot.
What about travelling with others?
Groups need slightly more structure, since undirected time tends to produce indecision rather than freedom. One agreed thing a day gives a group a spine without removing the rest.
What is the best use of an open day?
Returning to somewhere you liked, which almost nobody plans and many people find is the best part of a trip. A second visit to a neighbourhood, a market or a viewpoint without the pressure to see it is a different and better experience.
Does the length of trip change it?
Yes. A long weekend needs more fixing, since there is no time to recover from a wasted morning. A fortnight can absorb almost anything, which is why longer trips suit looser planning and short ones do not.
What about cancellation policies?
Where the flexibility is actually bought. A refundable accommodation rate costs modestly more and allows you to change your mind, which on a trip where you may want to stay somewhere longer is worth the premium.
What about the first day?
Worth leaving loose. Arrival days are unpredictable, with delays and tiredness, and a booking on the first afternoon is the one most likely to be missed. A walk and an early dinner is a better plan than a timed entry, and it orients you for the days that follow.
Which Multi-Country Routes Actually Work? goes into more detail on a closely related question, and the rest of our Inspiration section covers the wider subject.
The short answer
Fix accommodation, transport and anything with capped entry; leave the days themselves open. Keep a list rather than a schedule, build a buffer day before flying home, and use open days to return to somewhere you already liked.