How a plan row becomes individual posts with real proposed times — and why expansion never touches a week you have already laid out.
Jump to module
The content plan is the campaign's real planning object. The content mix you see as "3 images, 1 carousel" is a projection of it — each format's counts summed across the rows that name it.
Each row says four things:
Per-row accounts narrow the campaign's targeting; they never widen it. That is how you build a plan like "four image posts everywhere, one carousel on Instagram only" without splitting it into two campaigns.
The days field is the underused one. A row with no days is spread Monday to Friday. If you have a business where the weekend is the point — a restaurant, a gym, an events venue — naming the days is the difference between content that lands when people act and content that lands when the office is open.
A plan is a shape. A post prompt is one row of that shape made concrete: this post, this format, these accounts, this time. Expansion turns the first into the second, and the prompts are visible and editable on the campaign's detail page before anything is written.
Expansion answers exactly two questions.
Which weeks? Every week of the window that has not already been generated and that has no prompts yet.
Never a week that already has them. Not to add, not to remove, not to re-time. If you deleted four prompts from week three, you did that on purpose, and an expansion that helpfully put them back would be the software arguing with the person using it.
The practical consequence — and this is the surprise people hit — is that changing the plan changes the weeks that have not been laid out yet. Week one may already be written. That is also the only honest promise the system can make: it cannot retroactively unwrite content it has already produced.
So: get the plan right before you activate, and expect plan edits mid-campaign to take effect from the next unexpanded week onward.
Post prompts carry proposed times, and those times have real teeth — the draft saved for a prompt carries the prompt's time, and approving a draft keeps an existing future time rather than reallocating it. What you set here is what publishes.
The time is chosen in this order:
The layout is deterministic. Laying out week four does not depend on what happens to be in your calendar the moment somebody opens the page. Two people opening the same campaign see the same plan.
One exception is worth knowing: if a prompt's date has already passed by the time the run fires, the allocator takes over rather than writing a publish time in the past.
If your campaign content lands at times you dislike, the fix is the workspace preferred schedule, not the individual prompts. Editing prompts fixes one week; the schedule fixes every week and every routine as well.
A short set of rules that keeps prompt editing predictable:
Edit before activation where you can. Prompts for the first week are expanded early. Changes made before the first run is dispatched are free; changes made after apply from the next unexpanded week.
Deleting a prompt is a decision the system respects. It will not come back on the next expansion. If you want it back, add it yourself.
Prompts are per post, plan rows are per week. Do not try to express "one extra post this week only" by editing the plan — edit the prompts for that week. Do not try to express "one more image every week" by adding prompts — edit the plan row's count.
Times you set are promises the publisher keeps. If you move a prompt to Saturday 08:00, that is when it goes out, provided it is approved in time.
Check the accounts on each row before you activate. An account that names no valid connection is the single most common reason a campaign sits armed and generates nothing — the campaign will tell you so (Module 14), but it is easier to catch it here.
Reading a plan as a table is easier than reading it as prose. Each of these is one week; multiply by the duration for the total.
| Format | Count | Accounts | Days |
|---|---|---|---|
| Image posts | 3 | inherit | — |
Three images a week, spread Monday to Friday, on whatever the campaign targets. Nothing to get wrong. This is the right first plan for almost everybody.
| Format | Count | Accounts | Days |
|---|---|---|---|
| Image posts | 3 | inherit | — |
| Carousels | 1 | Instagram only | tue |
The images go everywhere the campaign targets. The carousel goes only where a carousel is worth making, on a fixed day so it always lands mid-week. This is the pattern that stops you paying for decks on platforms that render them badly.
| Format | Count | Accounts | Days |
|---|---|---|---|
| Image posts | 2 | inherit | thu, fri |
| Text posts | 1 | inherit | sat |
Named days, because for this business the weekend is the point. Without the days field all three would spread across Monday to Friday — which is the default, and would be wrong here.
Before activating, check four things: