Mental load app for couples: who plans ahead, not just who does it

The little one's booster shot is due before the end of the month. One of you noticed it, looked for a slot, booked the appointment and checked the vaccination record the night before. The other drove to the clinic. In most shared to-do apps, the task gets ticked by the person who drove. Everything else left no trace at all.

Most couple and chore apps divide up execution: who takes the bins out, who does the grocery run, who handles bath time. The mental load sits before and after the action itself: spotting a need, looking at the options, making the call, then making sure it actually got done.

In 2019 the sociologist Allison Daminger described this cognitive work, drawing on 70 in-depth interviews with members of 35 couples: anticipating needs, identifying options, deciding, and monitoring the result. In her study, women did more of this work overall, and more of the anticipation and monitoring in particular, while decision-making was roughly equal. The two most demanding phases were also the most unevenly shared.

Large time-use surveys don't capture any of this. In 2010, INSEE, France's national statistics office, measured 3 hours 52 minutes of domestic work a day for women and 2 hours 24 minutes for men in metropolitan France. Those figures time what gets done, not what gets thought about. Mental Loadless was built to measure that second share, between the two adults in a couple.

Inside the app, every chore and homework item records not only who carried it out but also who anticipated it, who decided and who checked. You don't fill in any extra fields: whoever creates a task is assumed to have anticipated it, whoever it is assigned to is assumed to carry it out, and a short weekly question, asked without any AI, confirms or corrects the picture.

From those attributions, the server calculates the MLI (Mental Load Index) every week: each adult's share of the household's load, on a scale from 0 to 100 where the shares add up to 100. The phases do not weigh the same:

Execution carries the lowest weight, on purpose. Anticipating weighs three times as much as executing, the reverse of an ordinary to-do list that only counts ticked boxes. A decision genuinely made together counts as zero on both sides: a shared call is nobody's burden. The method is versioned and published inside the app; the full weights and a worked example are on the page the science behind the MLI.

Alongside the MLI, an equity index looks at two phases only, anticipating and monitoring, and never at execution. It puts a number on a sentence many couples know by heart: “I do it all in my head, even when my partner is the one doing it.”

In many couples, the disagreement isn't really about the tasks themselves but about how each person perceives them. The crossed views screen puts those impressions next to the measurement.

Once a month, each adult separately gives two estimates: the share they think they carry, and the share they think their partner carries. The app then shows three figures for each person:

• What I think I carry — your own estimate
• What my partner thinks I carry — your partner's estimate of your share
• What is measured — the share calculated from the actual tasks

The gap between those three figures often says more than the measurement alone. It stops being one person's word against the other's: it becomes something between you that you can look at together without needing to win the argument.

A 30-second weekly check-in

Each adult can also answer a weekly check-in made of three sliders: rumination, energy, satisfaction. Answers are encrypted at rest and feed an eight-week curve shown next to the way your split evolves. What matters is the trend: whether a change in how you divide things is followed, a few weeks later, by a change in how each of you feels.

No, and the app never asks you to. A couple where one partner works nights, where the other is studying for an exam, or where a baby has just arrived has no reason to aim for an equal split every single week. What wears a couple down is rarely an imbalance they chose together; it is an imbalance that nobody ever decided on.

The distribution agreement lets the two of you set your own target. A deliberate 60/40 is a perfectly legitimate target. The app — and Coco along with it — then measures the gap to your target, not to some outside ideal. A 70/30 that you chose and kept to is an agreement honored.

Chosen imbalance

For particular stretches — a pregnancy, a crunch at work, a relative who needs care — you can declare a chosen imbalance: a period during which the uneven split is intentional. No reason is asked for. During that period, imbalance banners and rebalancing suggestions go quiet, Coco included. On the review date you picked yourselves, the app simply offers to take stock again.

The app's wording follows the same logic. Its texts use no word that sets up a grade, a winner or a party at fault, and no medical term to describe a person. The product's words are distribution and balance.

This is where rebalancing most often fails. You agree that your partner now handles the kids' medical appointments. They do handle them. And you keep thinking about it anyway: you glance at the calendar, you ask “did you call them?”, you keep the reminder on your own phone. The area has changed hands; the monitoring hasn't. We call this phantom monitoring.

The domain transfer is designed for that exact moment. It doesn't reassign a single task but an entire area of household life: the anticipation, the reminders and the checking. It works in four stages:

1. You choose the area and the adult who takes it over.
2. Reminders for that area are muted for the previous owner for 21 days, so they stop receiving the notifications that would pull them back into it.
3. Every Sunday, a reassurance message tells them the area is covered.
4. After the 21 days, the transfer is complete; it stays visible in the history and in the household's monthly report.

This three-week wind-down removes what keeps the checking reflex alive while trust settles in on the other side. And because the MLI counts monitoring as a phase in its own right, you can see in the following weeks whether the checking has really changed hands.

If you have never laid out who does what, the app offers a starting point inspired by Eve Rodsky's Fair Play method: 29 everyday tasks grouped into six categories — cooking, cleaning, children, admin, social life and health. Each task is weighted from 1 to 4 according to its effort, which is itself rated on five dimensions: physical, cognitive, duration, uncertainty and preparation.

You assign each task to one partner or the other, and the app works out the weighted split that results. For the method itself, its rules and its common pitfalls, read our article on the Fair Play method.

Fair Play and the MLI don't measure the same thing. Fair Play describes who is responsible for what, by agreement. The MLI describes who actually anticipated, decided and checked, week after week. The gap between the two — an area assigned to one partner but still anticipated by the other — is a concrete place to start the conversation.

Coco is more than a chat window: she acts inside the app. She adds or ticks off groceries, creates or reschedules an appointment, creates a task or a homework item, updates the pantry stock and reassigns a chore. “Move the dentist to 5 pm” reschedules the event. Any destructive action asks for confirmation first, and an undo button lets you reverse it afterwards.

For a bigger request — “organize Leah's birthday on Saturday” — Coco Architect drafts a plan of 2 to 8 actions (the event, the shopping, the invitations, the decorations), shown on a card you approve in one tap. A plan never deletes anything.

When your split drifts away from your target, two kinds of help are available. In the chat, ask Coco to rebalance: she suggests moving one to three unfinished tasks from the more loaded adult to the other one, and nothing changes until you approve it. The imbalance analysis, available from the Solo plan, suggests two or three concrete actions; it receives the load as anonymous ranks, never as first names. During a chosen imbalance, Coco doesn't bring up the split at all.

Coco runs on latest-generation AI hosted in Europe.

Yes, that is what they are for. Every month the app produces a household report: the split, the gap to your agreement, crossed views, check-ins, and domain transfers in progress or completed. The same content can be exported as a PDF for a professional, in nine languages, to put on the table in front of a couples therapist or a family mediator.

The PDF carries an explicit notice: the app measures and organizes; the professional interprets. It uses the same data as the report on your screen, and only your own household's data.

What the app doesn't do

• It doesn't replace couples therapy, or any professional support. It provides figures and neutral wording; what they mean in your relationship is yours to work out.
• It doesn't name a culprit. It describes a split, with no blame and no credit attached.
• It doesn't judge your target. 50/50, 60/40 or 70/30: that is your agreement to make.
• It doesn't live off your data. The free plan has no ads, and your data is hosted in France.

The free plan is enough to run the household together and sketch a first split with Fair Play. The measurement that compares two points of view needs the Family plan:

Solo only benefits the adult who subscribes. With Family, a single subscription covers both partners and the rest of the household. Every paid feature can be tried once for free; the details are on the pricing page.

Why does dividing household tasks create so much tension? Read our article on mental load in couples.