How do you split expenses when group membership changes mid-trip?

Tag every shared expense with who was actually present when it was incurred, not who's currently in the chat. Costs that span multiple membership windows (wifi, accommodation) pro-rate by weeks; costs that are point-in-time (meals, Ubers) split among presence-at-the-time only.

Can you split expenses across multiple currencies?

Yes, but capture the actual currency at the time of the expense rather than live-converting. At settle time, convert all expenses to a single base currency using the rate on the date each expense was incurred. TRIPTI.ai handles this with 38 currencies natively; a spreadsheet works too if you're willing to maintain the columns.

Should you pro-rate accommodation costs when someone joins late or leaves early?

Yes — split the accommodation cost by the share of the stay each person was present for. If the trip is 10 days and one person was there for 6 days, they pay 60% of the accommodation share they would otherwise have paid.

How do you handle the deposit on a co-living when membership changes?

The deposit is point-in-time, not a recurring cost. It's paid (and refunded) to whoever fronted it at the start. Don't try to split the deposit across people who joined later — it never applied to them.

How to Split Group Expenses When People Roll In and Out

Six weeks into the Lisbon co-living, you sit down to settle up. The first wifi bill spans weeks 1 through 3, when one set of people was there. The second covers weeks 4 through 6, when half the group rotated. Nobody can remember exactly who was in the apartment when, and the spreadsheet you started in week one has a column you don't understand. The math is fine. The membership window is the thing that broke.

In brief: When group membership changes mid-trip, standard expense trackers quietly fail because they assume the group is stable. The fix is presence-based tracking — tag every shared cost with who was actually there when it was incurred, pro-rate the costs that span membership windows, and pick a single base currency before you settle up.

The membership window is the thing that breaks

If you've ever co-lived with friends, organized a multi-week trip with a rotating cast, or extended a vacation solo while everyone else flew home, you already know this feeling. The math is fine. The problem is upstream of the math.

Standard expense apps were built for stable groups — eight friends at a beach house for four nights, everyone arrives Friday, everyone leaves Monday. Real long-stay travel doesn't look like that. Someone joins on day eight. Someone flies home on day twelve. Two people leave for a side trip to Madrid for the weekend. The wifi bill still arrives every month, the cleaning service still shows up every Friday, and you're left trying to reconstruct who was present for which expense from a chat thread that's now 4,000 messages deep.

The thing that breaks isn't the math — it's the membership window.

The thing that breaks isn't the math. It's the membership window.

The good news: the fix isn't complicated. It just requires changing what you're tracking. Stop tracking "the group." Start tracking who was actually present when each cost was incurred.

Why your current tools quietly fail

There's no single bad tool here. Each common option has a clean shape that almost fits, then breaks at the edges. Naming the gap is half the work.

Splitwise snapshots a group at creation. You set the members on day one, and the tool assumes that group is the group. When someone joins in week four, you have to either add them retroactively (and reassign every old expense), or create a parallel group that ignores the first three weeks. Either way, the audit trail gets messy and the planner spends an hour fixing it on Friday night.

Notion can work — but only if you manually add a "present for" column to every row. By week six, that column is the most-edited cell in the database, and the person maintaining it is starting to resent the maintenance. The system survives until the maintainer burns out.

Spreadsheets hold up for about three weeks, then collapse under the cognitive load. You start with clean columns. By week four, you have a "split-among" column that includes free-text names, a "currency" column with three different formats, and a comment thread on cell H47 you keep meaning to read. The spreadsheet didn't fail — it just stopped being readable.

WhatsApp loses receipts. The group chat is where the decisions happened, but it's not where the receipts live. Two weeks later, the "wait, who paid for the Tuesday groceries?" question has no answer.

None of these tools are wrong. They were built for a different shape of problem. They quietly fail at presence-tracking because they were never asked to track presence.

This is the kind of invisible coordination labor the planner ends up absorbing — not because they signed up for it, but because nobody else notices the gap until the spreadsheet melts.

Tag every shared cost with who was actually there

Here's the concept. Every shared expense gets tagged with the specific people who were present when it was incurred. Not who's currently in the chat. Not who's on the original guest list. Who was actually, physically there.

Three worked examples:

The wifi bill. Spans weeks 1 through 4. There were eight people in residence across that span, but not all at the same time. Tag the bill with each of those weeks, and split each week's share among whoever was present for that week. Person A who was there all four weeks pays a full quarter share. Person B who arrived in week three pays half of person A's share. The bill itself is one entry; the presence map handles the rest.

The Tuesday grocery run. Five of you went to the market, came home, cooked dinner together, ate it on the patio. Three other housemates were at a co-working day or out at their own dinner. Tag the grocery run with the five who shared the meal, not the eight people who happened to be in the lease. Point-in-time costs don't pro-rate.

The airport Uber. Two people, one ride to the airport, one expense. Tag it with just those two. The other six aren't in it at all.

The principle: a shared cost is shared among the people who actually shared it, not among the group that happened to exist on paper at the time.

How you structure this depends on the tool. In Notion, one row per expense plus a multi-select "present for" column. In a spreadsheet, one column per person, with a 0 or 1 for each row indicating presence. Either works. The principle survives the tool; the math is downstream.

This is the same shape of thinking that fixes the upstream coordination problem — the progressive scheduling method works by tracking what each person is actually committed to instead of demanding a single unanimous answer. Presence-based expense tracking does the same thing at the money layer.

Handle the multi-currency layer cleanly

Most multi-week trips cross currencies. Lisbon is EUR, the side trip to London is GBP, the airport tax went on someone's card in USD. If you try to live-convert every expense to a single currency as you go, you'll be wrong by the time you settle up, because exchange rates moved.

The fix is two steps:

Capture the actual currency at the time of the expense. The Tuesday grocery run was €42.30. Write down €42.30, not "around $46." You'll thank yourself in week eight.

Convert at settle time, using the rate on the date each expense was incurred. This is the only honest way to do it. The €42.30 from week one and the €38.10 from week five may not convert to the same dollar amount, even though the euro figures look comparable — exchange rates drift, and the person who paid in euros deserves the rate from the day they paid.

This is the part of co-living finance that most spreadsheets struggle with. TRIPTI.ai handles it natively with 38 currencies — capture the original currency on every expense and settle in whichever base currency the group prefers, with date-locked conversion underneath. If you're already living in a spreadsheet you trust, you can do the same thing manually with a few extra columns. The principle is what matters: don't live-convert.

When membership changes, don't restart — adjust

The instinct, when someone joins in week four, is to either ignore the first three weeks or restart the whole accounting. Both are wrong. There's a middle path that just requires distinguishing two kinds of costs.

Pro-rate the recurring costs. Accommodation, wifi, cleaning service, monthly co-working membership. These accrue over time, so the person who's present for half the period pays half the share they would have otherwise paid. The math is straightforward: split among everyone present during that cost's window, weighted by how much of the window each person was there.

Don't pro-rate the point-in-time costs. The deposit paid by whoever fronted it on day one. Individual meals. Optional excursions. These were single moments — they either applied to you or they didn't. The new person who joined in week four wasn't part of the day-one deposit, so they're not in the split.

The decision rule, in one sentence:

If everyone present when the cost was incurred shared the benefit, split among them. If the benefit extends across membership windows, pro-rate.

That's the whole framework. Everything else is just bookkeeping.

The fix is presence-based, not group-based

Most expense disasters at the end of a long stay are downstream of one assumption: that the group is the thing you split among. For long-stay travel, the group changes. The membership window is the thing that matters.

Tag every shared cost with who was actually there. Pro-rate the costs that span windows, split the point-in-time costs among presence-at-the-time only. Pick a base currency before you settle, and convert at the rate on the day each expense was incurred.

If you want this presence-based tracking built in rather than maintained by hand, TRIPTI.ai handles it natively — including multi-currency settle-up across the legs. But the principle is the principle: track who was actually there, not just who's in the chat.