A content calendar in a spreadsheet is one of those documents that starts full of good intentions and quietly stops matching what actually goes out. You put dates in, life happens, and by week three the sheet and the site have drifted apart. There is a way to get a month of posts on the calendar without keeping that document at all. You let topic hunts fill a pool, you prune it, and you let the project's own cadence do the placing.
Here is how that plays out in RankMill.
Start with a topic hunt
A topic hunt reads your website profile and comes back with five suggested topics. Each one has a title and an outline. You can start a hunt any time with "Find topics", and each project also runs its own on a cadence you pick: off, daily, weekly, every two weeks, or monthly, at an hour you choose. New projects start on weekly.
Five topics is not a whole month by itself, so if you want a bigger pool up front you can start another hunt once the first is done. Two hunts get you ten suggestions. There is a ceiling: a project holds at most 20 posts that are not out yet, suggestions included, and no hunt runs while it is full.
The point of doing it this way is that you are not sitting down to decide the shape of a month. You are looking at what came back and deciding what to keep.
Prune the pool, answer what matters
Keeping is the default. A suggestion sits in the project until something happens to it, and an unapproved suggestion costs nothing. So the work here is subtraction: decline the topics that do not fit. You can give a reason, and every later hunt is told your most recent declined titles and reasons, so it stops circling the same idea.
Some suggestions arrive with questions, up to four per topic: the things the topic finder could not find out on its own. A price. A customer name. A date. Whether you support some integration.
You can answer them. You do not have to. Questions are not a gate. A question left blank reaches the writer marked as not answered, and the writer is told to leave that fact out of the post rather than guess it. A piece on how you compare to a competitor, written with the pricing question blank, comes back without the pricing rather than with a made-up number in it. It is a slightly different post from the one you would have got by answering.
This matters when you want a month on the calendar in one sitting. Answering four questions each on ten topics is forty questions, and you are not going to do that. Answer the ones where the fact really matters and leave the rest.
Let the publish cadence place them
Now you have a pool. The next step in most tools is deciding what goes out on which day. In RankMill the project can do it for you, as long as it has somewhere to publish to: the cadence will not switch on until the project has a publishing destination.
Turn on the publish cadence, under the project's schedule settings, and pick three things. The weekdays posts go out on, in UTC. The hour the first post of the day goes out at. And how many posts each of those days takes, up to four, two hours apart. From then on the project checks every few minutes and keeps the next two weeks of slots filled with posts it already has.
Finished posts that are not on the calendar go first, because they are already written. After them come suggested topics, oldest first. You do not approve a topic to put it in the running. The cadence places it on a slot as it is, and writes it later:
- once its questions are answered, or straight away if it had none,
- or, if questions are still open, 24 hours before its slot, written without those facts.
That write is the moment credits are spent, 10 for the post, the same as approving it by hand. If the team is out of credits, the topic keeps its slot and the next check tries again. When the slot arrives, the post is published to the project's destination. On GitHub, that means RankMill opens a pull request and merges it.
Say you set Tuesday and Thursday, at 09:00 UTC, one post a day. The project takes the oldest candidate and puts it on the coming Tuesday, the next one on Thursday, and so on for two weeks. A day you have already filled by hand counts against that day's quota, so the cadence never doubles up on it.
You did not write a plan doc. You wrote a rule, and the rule placed the posts.
Two weeks, refilled
The window is two weeks at a time, not a month. A rolling fortnight is enough to see what is coming without pretending you know exactly what next month looks like. As days pass and new slots open at the far end, the project fills them from whatever is still in the pool.
The calendar shows the whole thing. Month view for a wider picture, week view for the next few days, filter to one project if a team has several. You can drag a planned post onto another day. A slot with no post on it just waits.
One thing catches people out. Clearing a post's time takes it off the calendar, but with the cadence on, an unscheduled post is back in the pool, and the next check may place it again. If you do not want a topic written at all, decline it.
When the pool runs low
At some point the candidates run out. This is where a plan doc would fall over, because you would need to sit down and pick new topics before the sheet made sense again.
When the cadence has slots it cannot fill, it starts a topic hunt of its own, on top of the project's regular hunts. Hunts are free. The new suggestions land in the project and become candidates for the next check. The extra hunt is skipped if one is already running or if the project is at its 20-post ceiling.
The opposite problem is just as real. A weekly hunt adds five topics a week, and two posts a week uses two. Leave that alone and the pool climbs to the ceiling, at which point hunts stop until posts go out. Declining the weak suggestions as they arrive keeps room for better ones.
If you want writing to start the moment topics arrive, there is a team setting that approves every new suggestion automatically, with a spending floor so the automatic writing stops while you still have credits left for the posts you ask for by hand. Those posts then sit in the pool as finished posts, which the cadence places first.
The Human in the loop switch
There is one setting worth knowing about before you turn any of this on. Per project, you can switch on Human in the loop. When it is on, a topic on the calendar with an unanswered question is not written until it is answered. There is no 24-hour fallback, only answered topics are placed on cadence slots, and a held topic whose slot passes unanswered gives the day back and stays a suggestion. It applies to topics you put on the calendar by hand too, so it matters even with the cadence off.
It is for projects where the facts really matter, and a post written around a missing number would not be good enough. Leaving it off is what lets unanswered topics flow onto the calendar without waiting for you.
What this looks like across a month
Put together, a month can go like this.
On the first you have ten minutes spare. The project already holds five suggestions from its weekly hunt, so you run one more and have ten. You decline two that miss, with a line each on why. You answer the questions on two topics where the facts matter, and leave the rest.
Your publish cadence is set to Tuesday and Thursday, one post at 09:00 UTC. The calendar shows the next four slots filled. Each topic is written when its answers land or a day before its slot, and goes out on its day.
The weekly hunt keeps adding suggestions. Once a week you skim them and decline what does not fit. Two posts a week is at most ten in a month, which at 10 credits each fits inside the default allowance of 100.
You did not open a spreadsheet once.
When to still write by hand
None of this stops you writing a post from your own idea. "Write from an idea" takes a title and a short brief, and the writer starts straight away. You can also import a post you already have and keep it as is, restructure it, or rewrite it in your project's voice.
Once those posts are finished, they are candidates like any other. A finished post that is not on the calendar goes ahead of the suggested topics when the cadence fills a slot.
Why this works
The reason a plan doc goes stale is that it is trying to be two things at once: a list of ideas and a schedule. When one moves, both are out of date. Splitting them apart, so the pool of posts is one thing and the cadence that places them is another, means neither goes stale on its own. Topics arrive on their own timer. The cadence uses them when the days come around. If the pool runs dry, the cadence goes looking. If you stop the cadence, the posts sit and wait.
You never write a month down. You write one rule, and prune.