Why I built Travel Wand

The trip is real. We are planning a loop through Kenya and Tanzania, a flight south, and a stretch of South Africa at the end. Somewhere around the tenth revision of the spreadsheet, I admitted that the tool was the problem and started building the one I wanted.

Spreadsheets are the natural tool

Nearly every trip I have planned started as a spreadsheet, for good reason. Days become rows, costs sum themselves, and everyone can read it.

It breaks down in three ways. The first is distance. A spreadsheet has no sense of geography, so two rows can sit next to each other while the places they name are six hours of driving apart. You usually discover this later, while checking something unrelated on a map.

The second is detail. A trip is full of small items: travel times, rest afternoons, restaurants you want to try. Add them all as rows and the plan becomes hard to read, because the rows that actually shape the trip, the attractions and the tours, get buried in the filler. Leave them out and the plan quietly lies about how full each day is. There is no good level of detail.

The third is change. A travel plan is mostly approximations, and a spreadsheet treats every cell as certain. When the dates of a trip shift by two days, something has to give, and the spreadsheet cannot tell you what. It does not know which bookings are fixed, which days can compress, or which drive can happen one day later. You end up re-deriving the whole plan in your head every time anything moves.

The map knows where, and not much else

Google Maps is genuinely great at the other half of the problem. It keeps track of every location, the saved pins never get lost, and it answers the distance question instantly.

It has no concept of time. There is no timeline, no ordering, no notion of an itinerary that spans days and cities. A list of starred places tells you nothing about which day any of them belongs to. For finding one place the interface is excellent. For planning a whole trip it is good, and it never quite becomes great.

The research never lives in one place

There is a third problem that neither tool touches. The important points about a place are scattered across the web. The ferry timetable lives in a blog post from 2019, the advice about which entrance to use lives in a forum thread, and the one review that mentions the early morning light is buried on page four. Planning well means collecting all of it, and by the end it lives in twenty open tabs and a notes file you will never find again.

The places worth traveling for

The part of travel I care about most is the part no tool has ever helped with. In every new city I look for the hole-in-the-wall places: the counter with four stools and one perfect dish, the walk that only makes sense at a particular hour. Finding them takes exactly the kind of scattered research described above, and they are the first thing to fall out of a spreadsheet because they look like clutter next to the big-ticket rows.

Those finds are usually what we remember a trip by. A planning tool should give them a place to live.

So I built Travel Wand

Travel Wand keeps the map and the timeline on the same screen. The map shows what is possible, the plan shows what you have chosen, and the two stay in sync as details change. Each place carries its own description, costs, and sources, so the research has one home, and the small finds sit in the catalog next to the famous ones. When dates move, you can branch the plan, try the change, and compare the two versions side by side.

The first trip it plans is our own. I will use this journal to write about building it, including the choices that work and the ones that need another pass.