A content calendar holds up when it has two layers, a fortnight locked piece by piece and a month sketched by topic, a buffer of three pieces already shot and a written rule about what gets dropped first. What breaks a calendar is never a shortage of ideas. It is that nobody decided in advance what happens when one piece falls through.
Where it always breaks, and it is not where you think
Nobody abandons a calendar because they ran out of ideas. It gets abandoned on the Tuesday when Wednesday's piece depends on a photo that has not arrived, there is nothing behind it, and the seven in the evening decision is to publish something worse or publish nothing. That decision always goes badly, because it is made in a hurry, and the following week starts already behind.
Three failures repeat across every account. The first is the outside dependency: a piece that needs a file, a figure or a sign off from somebody who is not looking at your calendar. The second is the month with two bank holidays and a week off, which on a spreadsheet looks like an ordinary month and in practice has a third fewer working days. The third is the approval that lands after the hour, when the piece has already lost its slot.
None of the three is fixed by planning further ahead. They are fixed by planning in states rather than in dates, which is what separates a calendar from a list of good intentions with checkboxes next to them.
The four states a piece can be in
A sheet of dates and headlines cannot answer the only question that matters on a Tuesday: out of everything in this fortnight, what is actually ready. To answer it, every piece has to sit in one of four states and in exactly one.
| State | What it means precisely | Who moves it | How long it can sit there |
|---|---|---|---|
| Idea | A topic and a pillar, no script and no firm date | Whoever owns the editorial line | Until its fortnight begins |
| In production | Somebody is filming, writing or editing it | Whoever produces | Days; if it is weeks, it is blocked |
| Ready | Final file, caption written, alt text, network chosen | Whoever produces | Hours, or it goes to the buffer |
| Scheduled | Date, time and account assigned, waiting its turn | The calendar | Until its hour |
The point of states is not the order, it is that they make the jam visible. If Monday finds six pieces in production and none ready, the fortnight is already lost even though the calendar still paints seven green slots. A spreadsheet will not warn you about that; a board by state will, and that is the only real reason to change tool.
There is a fifth state worth naming even though it is not quite a state: blocked. A piece that has been waiting five days on something external is not in production, it is stopped, and mixing the two hides the jam until it is too late to do anything about it.
Two layers: the locked fortnight and the sketched month
Locked does not mean set in stone, it means nothing is left to decide. A locked piece has the final file, the caption written for the narrowest network in the group, the alt text, the network or networks chosen and the hour set. If any of that is outstanding, the piece is not locked however firm its date looks.
Sketched is the opposite and it also has a definition: pillar, format and week. No caption, no script, no exact date. The urge to write a whole month of captions in one sitting feels like productivity and is waste, because half of those pieces will change before their week arrives and you will have written for the bin.
The split between pillars is what lets a sketched month be written in twenty minutes: once it is settled that half the month teaches something, a quarter shows the work and a quarter sells, the grid fills itself from there. How those pillars are chosen and reviewed has its own content pillars guide. And the number of slots you draw on the grid is not an aspiration, it is the result of how many hours you have and what a finished piece costs, which is the arithmetic in the posting frequency guide.
If you would rather start from a grid that is already laid out, the content calendar builder spreads the month's slots across your cadence and your pillars and hands it back as a CSV. It lays out slots and nothing else: what goes inside each one you write yourself, here as anywhere.
The buffer, and the written rule about what falls
Three pieces shot, with no date and no expiry. That is the whole mechanism, and it is the difference between a bad month and a broken one. They are not filler: they are the ones you liked that had no slot, kept in the ready state and not in the idea state, which is the distinction almost everybody gets wrong. A folder of ideas does not rescue a Wednesday. A finished file does.
Not every piece belongs in the buffer. It has to be one that does not go off: nothing tied to a date, to a promotion that ends, to a figure that ages or to a topical reference nobody will follow in three weeks. What works best is whatever answers a question you get asked all year, because you will be asked it again next month. Shooting two of those while the kit is already set up costs twenty minutes and buys a whole Wednesday.
The buffer is refilled in the same session it is spent, and if it goes unrefilled for two fortnights running, the problem is not the buffer, it is that the cadence sits above what you actually produce. The answer there is to lower the number, not to push harder.
The second half of the mechanism is a three line rule written before you need it, because nobody decides well on the 12th. The first thing to go is the number of networks, not the number of pieces: holding three posts on one network survives better than letting one a week slip on five. The second is the expensive formats, so a ten image carousel becomes one good image. The last thing you touch is the caption, the only item on the list that costs no money and does decide whether anybody stays.
Everything in this article happens in one place in GoFeed.
Try it freeWhen somebody else has to say yes
Your own account gets scheduled and that is that. The moment there is a team, or a client who reviews, a step appears between ready and scheduled that eats calendar days and that almost nobody draws. The practical consequence is that production has to run a week ahead of publication rather than two days, and that the date written into the proposal is the publication date and not the delivery date.
It is worth being exact about how that step works, because half the calendars out there oversize it. Your team or the client approves from their portal: one yes is enough. And for a client who doesn't need to review, posts are scheduled straight away. It is one approval and there is no order to who gives it, and a piece whose hour passes while it waits does not go out late: it sits waiting for somebody to give it a new date.
That is also the point where a spreadsheet stops working, not because of the number of pieces but because of the number of hands. Two people editing the same row is a broken sheet, and a client who has to be resent the sheet every Monday is a sheet nobody opens. The publishing calendar exists for that specific reason, and what it does with one piece and several networks is on the calendar and publishing page.
The twenty minute review
A calendar is maintained on Mondays, in a short review that is always the same. What went out last week and what did not. What is blocked and waiting on whom. How many pieces are in the buffer. What is locked for the fortnight starting now and what the rest is missing.
Four questions, twenty minutes, and most weeks they are answered in five. The value is not in any single week's answers, it is that by the third month you can see the pattern: the same kind of piece always jams, the same person is always the one missing, the same day always slips. That is not fixed with more planning. It is fixed by changing who does what, and it is the only genuinely useful conclusion a calendar can hand you.