Groups rarely fail at cooking. They fail at the twenty minutes before cooking: who decided this, who bought the coriander, whose turn is it, and does anyone remember that one of us stopped eating pork in March. Every one of those is a coordination problem, and coordination problems are solved by making the plan visible and the responsibilities named — not by trying harder.
Here is a routine that works for a family, a shared flat, a couple, or a group of friends with a standing Thursday dinner.
Step 1 — Create the group and be honest about roles
Set up a group and invite people with the join code. Roles are worth a moment's thought rather than making everyone an admin by reflex:
| Role | Good for |
|---|---|
| Admin | The people who actually maintain the plan and finalize the week |
| Participant | Everyone who eats and should get a vote and a cooking night |
| Viewer | People who need to see the plan without changing it |
Anyone with allergies should share their allergen profile with the group at this point. It is opt-in on purpose — sharing a medical detail with a group should be a decision, not a default — and once shared, every dish proposed for that group's plans is checked against it.
Step 2 — Draft the week, badly and fast
One person drafts. This is the step groups get wrong most often: they try to decide together from an empty grid, which is the slowest possible way to fill fourteen slots. A draft is much easier to react to than a blank page, so somebody should fill the week quickly — from the shared rotation, or by generating it — knowing full well that half of it will change.
Two rules make the draft good enough to react to. Match effort to the calendar, so the late nights get the fast dishes. And leave at least one slot deliberately empty for the night that always goes sideways.
Tip: draft from what the group has already liked. Filter the recipe list by rating, or by the dishes that have been in a plan before, and you are choosing from a shortlist your group has effectively already voted on.
Step 3 — Let people vote, with a deadline
While the plan is a draft, members approve or reject the proposed dishes, and can leave comments on the plan itself rather than in a chat thread that scrolls away. Give the vote a deadline — "by Saturday evening" — because an open-ended vote is how a plan stays a draft until Wednesday.
Treat rejections as information rather than a veto: two rejections on Thursday's fish curry means swap Thursday, not reopen the whole week. The point of voting is to find the two dishes nobody wants, not to reach consensus on all fourteen.
Step 4 — Finalize and name the cooks
An admin finalizes the week. Now assign cooking nights — by name, in the plan, not by an understanding that Tuesday is "usually" someone's. Owning a night means owning the whole night: cooking and the clean-up. Groups that split the tasks within a night (one cooks, one washes) generate more negotiation, not less.
If someone is travelling, the plan should say so. An unassigned night is honest; a night assigned to someone who is on a train is how a plan collapses.
Step 5 — One shop, one list
Generate the shopping list from the finalized week. Every ingredient scaled to the servings in each slot, duplicates merged, grouped by aisle. Whoever passes a supermarket takes it — and because the list is shared and ticked off as it is bought, two people don't come home with the same chicken.
Check the pantry before you go so the staples you already own don't get bought again, and add what you bought to the pantry when you unpack. Five minutes of bookkeeping, and next week's list starts from what is genuinely in the house.
Step 6 — Rate what you ate
This is the step everyone skips and the one that compounds. After the week is cooked, rate the dishes. Group ratings turn the shared rotation into a ranked list, so next month's draft starts from what this group actually enjoyed — and a dish everyone quietly disliked stops silently reappearing every three weeks because one person keeps proposing it.
Over a couple of months this is the difference between a group that plans and a group that plans well: the draft gets faster, the votes get shorter, and the rotation converges on food this particular set of people likes.
Four ways group planning goes wrong
- Planning by committee from an empty grid. Draft first, react second.
- Everyone is an admin. If nobody owns finalizing, the week never gets finalized.
- Planning all seven nights. Most groups genuinely share three or four. Plan those; leave the rest alone.
- No named cook. "We'll see who's free" is how a planned night becomes a takeaway night.
Frequently asked questions
How long should this take each week?
Ten minutes for the draft, a couple of minutes each for the votes, five for the list. The routine should be shorter than one round of "what does everyone want".
What if someone never votes?
Finalize anyway, on the deadline. A vote is an opportunity, not a requirement — and people who don't use it usually turn out not to mind.
Can guests who don't use the app be included?
For a one-off dinner or party, use an event with an RSVP link: guests reply and send their dietary needs without installing anything, and you plan the menu against what comes back.
Can one person belong to several groups?
Yes — a household group and a Thursday-dinner group with friends can run side by side, each with its own plan.
Cook together without the coordination tax
Share one plan, vote on the draft, assign the nights, and shop once. Mazir handles the bookkeeping so the group only has to make decisions. Free to start.
Try Mazir