Nobody has ever been angry about a plate
Think back to the last time it went wrong. One of you said something about a task. The other heard something about a character. Two minutes later the pan was still in the sink and you were arguing about last summer.
That happens because a household task has two halves and only one is visible. There is doing the thing, and there is knowing the thing needs doing. The second half runs in the background, produces nothing you can point at when it goes well, and in most homes it sits on one person far more than the doing does.
This is why asking counts as work. “Just tell me what needs doing” is almost always said in good faith, and it hands the second half straight back. The person who asks is still holding the schedule, still picking the moment, still finding a phrasing that does not sound like an order.
Asking is not a shortcut around the work. It is the work, said out loud.
The other side is just as real and usually goes unsaid. The person being asked is not, in most homes, refusing. They are running a different set of trigger points. One of you sees a sink that has to be empty before bed, the other a sink that can wait until after the film. Both readings are sincere, both are partial, and that is why the conversation goes in circles.
Why 50/50 fails as a target
50/50 is attractive because it sounds checkable. It fails because a task count treats every line as one unit, and household lines are wildly unequal units.
Taking the bins out is a few minutes once a week, on a deadline set by the truck. Cooking on weeknights is five decisions and five sittings, each needed before anyone is hungry. On a list, both are one line, and a count says they cancel out.
Three things make two lines unequal, and duration is only one of them.
| Line | Comes back | Costs | Who has to notice |
|---|---|---|---|
| Bins | Weekly, on a fixed day | Minutes | The truck timetable |
| Weeknight dinners | Five times a week | An hour, plus the decision | Whoever gets hungry first |
| Laundry | When the basket fills | Spread across a day | Whoever runs out of clothes |
| Dentist, vet, insurance | Once or twice a year | A phone call | Always the same person |
Four lines, and a task count splits them two each. Anyone who has lived with another person knows that is not a split.
A better unit is the cycle. Ask how often a line returns, whether it can be moved when a week goes wrong, and who has to notice it is due. Then hand over whole cycles rather than turns, and record the noticing as its own column, because that column decides how the evening feels.
Three steps, one evening
1. Write down everything that comes back
Not the chores you remember, everything that returns. Include the lines that are pure attention: refills, appointments, the booking that costs double if it slips past September. Leave one-off jobs out, because a dripping tap happens once and then stops costing anything.
Do it across a normal week rather than from memory on a Sunday night. Memory is the thing that got you here. More on that step, including how to record the noticing itself, is in making the mental load visible.
Expect the sheet to be longer than either of you guessed. That length is not an accusation, it is the reason you disagreed.
2. Agree what done means, line by line
A lot of chore friction is not about who does it. It is about what counts as finished. If nobody has said that out loud, the person with the higher standard ends up inspecting and the other one feels supervised.
Write the definition in the flattest language you can manage. Kitchen done means sink empty, worktop wiped, hob wiped. Bathroom done means basin, mirror, floor, bin emptied.
Then the rule that makes it hold: nobody re-inspects and nobody silently redoes. If the agreed definition is too low for one of you, raise it out loud and see whether the other signs up. Redoing someone’s work without saying so is the surest way to make them stop.
3. Assign by cycle, then fix the date you revisit it
Hand over whole cycles. One of you owns the weeknight dinners, the other the bathroom and the bins. Trade on genuine aversion where your aversions do not overlap, because that trade is free. Preference will not cover the whole list, and the lines nobody wants rotate on a period long enough to notice, months rather than days.
Then pick the review date while nobody is angry. Fifteen minutes, once a month, in the calendar. People skip this step and it is the one doing the work, because the alternative is reopening the split at the moment somebody snaps. Nothing agreed then survives the week.
Two households
Two schedules that do not overlap. One of you works shifts, the other has fixed hours. The fixed-hours person carries every weekday evening and starts to feel like the default. The one on shifts does two large blocks on a day off, sees the flat transformed and cannot understand the resentment. Both accounts are accurate. Splitting by cycle eases it: the weekday loops go to whoever is reliably there on weekdays, the long blocks to whoever has whole days free, and the sheet says so instead of it being assumed every Tuesday.
Two standards in one flat. You moved in together recently. One of you lived alone for years and has a working system that is genuinely more demanding. The other managed fine with a lighter one. Here the split is not the problem, the definition of done is. Write the definitions first, negotiate them once, then let each line be finished by whoever owns it. If one of you cannot live with the agreed standard on a line, that line belongs to them at their standard, not to both of you at two.
When the split really is unequal
Sometimes the sheet does not reveal a misunderstanding. It shows that one person is carrying most of the house, and that is simply true.
Say this plainly: a written list makes an imbalance visible and precise, and that is all it does. It does not produce the will to change it. If one of you reads the finished sheet and does not accept what it shows, you are no longer in a scheduling problem. You are in a conversation about what each of you thinks is owed to the other, and that one is out of reach of any app, ours included.
The reverse happens too. The sheet sometimes shows that the split was closer than one of you felt, and that the weight sat in one season rather than spread across the year. That is also a result, and it also has to be accepted.
What keeps either version survivable is treating the sheet as evidence about lines, not a verdict about people. A line with no owner is fixable. A partner who does not care is not a claim a list can support.
When the list becomes a weapon
It can, and you will recognise it quickly. Someone opens the shared list mid-argument to prove a point. Lines get added after the fact so the tally leans. Someone ticks a line off on the other person’s behalf, loudly. Or someone stops ticking anything at all, the quietest way to leave the table.
Three rules keep it a tool.
- The list is read at the review, not during a disagreement. If you need it to win an argument, it has stopped working for the house.
- New lines are said out loud before they go in. A list that grows silently is no longer shared.
- The list is not a history. What matters is the next occurrence, not who missed what in March.
That last one is also a product decision.
What we built, and what we refused to build
This article comes from the people who make Bunkly, an app for shared household tasks. This is the section where we have an interest, so it is signposted.
There are no points, no badges, no leaderboard and no mascot, and that is a refusal rather than a missing feature. In a flatshare a scoreboard is a nudge. In a couple it is evidence. Once a number exists, any disagreement can be settled by opening it, and whoever stops playing first looks like the one who stopped caring. A household has to be able to forget a bad fortnight. A score never does.
Two other decisions come straight from the problem above. Recurrence is anchored on the day a task was actually completed, not on a fixed weekday, so one missed occurrence does not cascade into a wall of overdue lines, and that wall is an argument waiting to happen. On the paid plan, rotation between members assigns itself, so neither of you has to be the one who reminds the other, and that reminding is usually the real grievance.
One honest limit. Undoing a completion recomputes the next due date from the previous real completion, which is what you want after a mis-tap. If you had moved that due date by hand before the mis-tap, the undo restores the calculated date, not your manual one.
The free plan is a complete single-person plan, which matters here more than it sounds. One of you can write the whole list alone, across a normal week, and bring something finished to the table instead of an empty screen.
What it does not do. It does not decide what is fair, does not weigh one cycle against another, and cannot tell you whether your split is just. It will not hold the conversation two sections above for you. There is no shopping list, no calendar, no meal planning and no web version. The features page lists what is in, and the pricing page has the free plan, the subscriptions and the one-off purchase.
What fair means here
Not an equal count. A split both of you would still sign at the end of a bad week, written where neither of you has to remember it, with a date for changing it when the bad weeks stop being exceptional.
Start with the sheet. Do the boring part first, because it is what lets you argue about a line instead of about each other.