A meal plan that survives a real week
Most meal plans die on Wednesday. The fix isn't more discipline: it's planning fewer meals, in no fixed order, with the week's chaos already built in.

Almost everyone has tried meal planning. Almost everyone has stopped.
The pattern is consistent enough to be funny. Sunday: a full grid, seven dinners, a tidy shopping list, a genuine sense of having got your life together. Monday and Tuesday go beautifully. Wednesday someone works late. Thursday you do not want the thing Thursday-you was assigned by Sunday-you. By Friday there is a bag of spinach turning to liquid in the drawer and a quiet decision never to do this again.
The usual diagnosis is that you lack discipline. That is wrong, and it is why the next attempt fails the same way. The plan failed because it was a schedule, and your week is not a schedule.
Seven dinners, one per day
- Three meals that were never going to happen
- Food bought for them anyway
- A Wednesday that makes you the one who failed
Four dinners, no days attached
- One shop, four decisions already made
- Tonight is whichever of the four fits tonight
- Nothing to fall behind on
Here is what works instead.
Plan four dinners, not seven
The single highest-value change, and the one people resist most.
You do not eat seven cooked dinners a week. Nobody does. There is a night you eat out, a night of leftovers, a night that turns into toast at half nine because of something that was not in anyone's diary. Planning seven means planning three meals that will not happen, buying food for them, and then feeling like a failure when reality arrives on schedule.
Plan four. Shop for four. The other nights take care of themselves the way they always have, and (this is the part that surprises people) you stop wasting food almost immediately, because you stopped buying for meals that were never going to be cooked.
If four feels like too few, you can always cook something extra. The failure mode you are designing against is not "not enough plan." It is a plan with slack in it that you can actually keep.
Do not assign days
The second change, and it costs nothing.
A meal plan should be a shortlist, not a timetable. Four dinners you have decided on, the shopping done, and then each evening you pick whichever of the four suits the night you are actually having.
Assigning Wednesday to the thirty-minute traybake means that on Wednesday, when you get home late and exhausted, you are looking at a task rather than a choice, and the natural response to a task you did not choose is to order a takeaway.
The same four meals, unassigned, survive contact with a bad Wednesday. Tired? Take the quick one. Weirdly energetic? Take the long one. Nothing has gone wrong, because nothing was scheduled to go wrong.
This week





Build the week around one long cook
Every good week I have seen has one meal doing double duty.
Cook one thing properly at the weekend (a ragù, a roast chicken, a pot of beans, a big tray of roast vegetables) and design two of your four weeknight meals to start from its remains. Roast chicken becomes a stock and then a soup. Ragù becomes pasta and then a baked thing with cheese on top. Beans become a stew and then something on toast with an egg.
This is not the joyless "batch cook nine identical containers on Sunday" version, which nobody sustains because eating the same meal five times is grim. It is one long, enjoyable cook that leaves you with a component, plus two short weeknight meals that are genuinely different dishes.
Shop from the plan, not from the shop
The shopping list is where most plans quietly leak money and food.
Two rules. Check what you already own before you write the list, the reason there are three jars of cumin and no eggs is that nobody checked. And buy for the plan, not for the feeling of being well-stocked. The vegetables you buy because they look good, without a meal attached, are the vegetables you throw away.
This is the bit worth automating, because it is pure bookkeeping. In Sugo, a meal plan rolls up into one shopping list, deduplicated across the recipes, with the things already in your pantry taken off it. That list is also the thing that makes the four-meal plan work in practice: one shop, four dinners, no midweek top-up.
Shopping list
- Pecorino200 g
- Spaghetti500 g
- Gnocchi400 g
- Kale1 bunch
- Smoked bacon150 g
- Black pepper
- Coconut milk400 ml
- Olive oilrunning low
Let the plan change
A plan that cannot change is not a plan, it is a promise, and promises to your future self are famously unreliable.
Swap freely. Move a meal to next week. Delete one because you have gone off the idea. If two of the four get cooked and the other two slide, that is a successful week, not a failed one. You cooked twice from a decision you made once, which is exactly the point.
Cooking for a household makes this more important rather than less. A plan that lives in one person's head is a plan that person has to defend every evening. A shared one that everybody can see and change is just information, and the negotiation about what to eat happens once on Sunday instead of every night at seven.
What a workable week looks like
Four dinners
Chosen from recipes you already have and have already liked.No days attached
Pick each evening from the four.One long weekend cook
Something that feeds two of them.One shop
From a list that knows what is already in the cupboard.Permission to bin the whole thing on Thursday
Without it counting as failure.
That is a meal plan that survives a real week: less ambitious than the grid, and the only kind anyone is still doing in March.
Sugo is in open beta on iPhone. If you have abandoned meal planning more than once, join the beta.