Every ecommerce founder we work with can tell us, from memory, how many weeks of cover they have on their hero SKU. Not one of them, on day one, can tell us how many weeks of content cover they have. The feed going quiet in November was visible in September, in a number nobody was looking at.
This is the supply metric for a content system. It sits upstream of everything we have written about banks, routers and running orders, and it is the one that tells you a stall is coming before the stall.
The stall is never a writing problem
When a founder's publishing drops from three posts a week to one, the diagnosis in the room is always about the writer, the schedule, or motivation. In our experience it is almost never any of those. The banks ran dry, and nobody noticed because nobody counts what goes into them.
A content system consumes raw material. A proof post consumes a proof-bank entry. A mechanism post consumes an explanation-bank entry. A story post consumes a story. Those entries do not regenerate on their own. They arrive when the founder is close to the live work and captures what they see, and they stop arriving when the founder gets busy, gets distant, or has a bad quarter.
The consumption side is steady: two or three posts a week, every week. The supply side is lumpy. That mismatch is the whole problem, and it is exactly the shape of an inventory problem, which is why founders understand it instantly once it is framed that way.
What content days of cover actually measures
The metric is simple enough to run in a spreadsheet you already have.
Usable entries in the banks divided by weekly consumption, expressed in weeks.
Usable means an entry that could become a post this week without any further work from the founder. A screenshot with no note is not usable. A number with no window attached is not usable. A story with a live counterparty is not usable. Count only what would clear your own pre-publish checklist today.
Consumption is the number of bank-backed posts you ship per week. Not total posts. Opinion posts and position posts do not draw down a bank, so they do not count against cover. For most of our clients it is two to three.
A founder with 14 usable entries shipping three a week has under five weeks of cover. That founder feels fine, because five weeks is a long way off and the calendar is full. Five weeks is not a long way off. It is one busy month.
We track it per bank, not blended. Proof cover, story cover, explanation cover. A founder can have twelve weeks of explanation cover and one week of proof cover, and the blended number will say seven and hide the fact that the posts which actually generate inbound are about to stop.
Why the trend matters more than the number
The level tells you where you are. The trend tells you what is coming.
Cover falling for three consecutive weeks means supply has dropped below consumption. The founder has stopped capturing. That is the earliest signal we have found for a content stall, and it shows up 30 to 60 days before the feed goes quiet, because the banks buffer the gap until they don't.
Every stall we have post-mortemed had the same shape. Capture slowed in month X. Nothing visible changed in month X because the banks were full. Publishing thinned in month X plus two. The founder noticed in month X plus three, blamed the writer, and by then the archive had a hole in it dated to the exact quarter things got busy.
The trend is what you act on. The level is what you report.
How the banks drain, in order
The banks do not empty evenly, and the order they empty in is diagnostic.
Proof drains first. Proof requires a result, a result requires time, and a founder under pressure stops running the tests that produce results. This is why a founder in a difficult quarter feels like they have nothing to say when they have more raw material than ever. The proof bank is empty. The mechanism bank is full and they are not looking at it.
Story drains second. Stories are captured in the moment or not at all. A founder who has moved off the tools stops generating them. A founder who is on the tools but stopped capturing loses them at the same rate.
Explanation drains last. Explanations are pattern-level. They survive distance from the live work. They also produce the posts founders find least exciting and readers find most useful, which is why an account running on explanation cover alone reads more general than it used to, even though nothing about the quality dropped.
If proof cover is under three weeks while explanation cover is over ten, the founder has not run out of things to say. They have run out of receipts, which is a different problem with a different fix.
Restocking is scheduled, not hoped for
Once you can see the number, the response is the same one an operator uses for a SKU: reorder against a lead time, before you hit the floor.
Set a floor per bank. Four weeks is where we start alarming on proof. Six on story. Explanation can run lower because it refills fastest.
Schedule the restock as a dated deliverable. Not "let's find some more proof." A named sync, on a date, whose only job is to pull ten entries out of the founder's last month. A restock sync is a different conversation from a normal voice sync. The question is not "what's new," it is "what did you check, change, or kill in the last thirty days, and what number moved."
Know the lead time. Proof has the longest lead time because a result has to happen before it can be captured. If proof cover is four weeks and a test takes six, you are already short. Start the test now for the post in eight weeks.
Restock the bank that is short, not the one that is easy. The instinct under pressure is to fill whichever bank fills fastest, which is always explanation. That produces a founder with twenty weeks of explanation cover, no proof, and a feed that has quietly gone general.
Days of cover versus the pipeline board
We run a pipeline board with every client and it is a different instrument. The board tracks drafts through stages, so you can see where production is bottlenecked this week. Days of cover tracks raw material before it becomes a draft, so you can see whether there will be anything to put on the board next month.
A full board with empty banks looks healthy for exactly as long as the board takes to clear. Then it is empty and the reason it is empty was visible six weeks earlier in a number nobody kept.
The board tells you the writer is slow. Cover tells you the founder stopped capturing. Those get fixed by different people.
The founder-side version
Not every founder wants a spreadsheet. The version that survives contact with a busy quarter is one question at the top of every sync, answered from the bank not from memory:
How many posts could we ship next week without you giving us anything new?
If the answer is fewer than the number of posts you publish in two weeks, you are in the window where a stall is being decided. The feed still looks fine. The archive is about to acquire a gap with a date on it, and the fix costs one conversation now and a quarter of rebuilding later.
FAQ
Is this just a content backlog count? No. A backlog counts drafts and ideas. Cover counts usable raw material against a consumption rate, per bank, as a trend. A backlog of forty ideas with no proof behind any of them is zero weeks of proof cover.
How often should it be measured? Weekly, at the point you fill the running order. It takes two minutes once the banks are maintained, and the trend is only readable if the interval is consistent.
What if the founder has genuinely run out of material? Then the number told you a month earlier than the feed would have, and the conversation is about supply rather than about the writer. Usually the founder has not run out. Usually one bank is empty and two are full.
Does this apply to a founder posting once a week? More. At one post a week, every bank-backed post is a larger share of the annual archive, and a four-week gap is a visible hole. Consumption is lower so cover is longer, which makes the trend easier to ignore and the stall more expensive.
Who owns the number? We do. The founder owns the supply. The one input we cannot generate is what happened in their business this month, and the metric exists to make sure we ask for it before we need it rather than after.
If you want a content operation that reports on its own supply instead of discovering a stall in the feed, that is the system we build with every EcomGhosts client.