The engagements that fall apart rarely fall apart on content quality. They fall apart in August, September and October, when the founder's operating calendar quietly eats the one input nobody can buy.
We've now run enough ecommerce ghostwriting engagements to see the shape of it. A client who was excellent for five months goes thin in a specific window. The voice syncs get shorter. The stories stop arriving. Approvals come back in four minutes with no notes. And then the cadence slips, and then it slips again, and by the time anyone calls it a problem the founder has been effectively absent for six weeks.
It gets diagnosed as disengagement. It's almost never disengagement. It's a bandwidth collision, and for ecommerce founders it is the most forecastable failure in the entire relationship.
Ghostwriting has two inputs and only one of them is purchasable
This is the structural fact underneath everything else.
The first input is writing time. That's ours. It's schedulable, it's staffed, it doesn't care what week of the year it is, and it's the thing the invoice is nominally for.
The second input is raw material — the story about the container that sat in Long Beach, the actual return reason codes, the number the founder pulled on Tuesday, the argument they had with a supplier. That input does not come from a calendar slot. It comes out of the same hours the business is consuming.
So when the business gets loud, the writing capacity stays exactly where it was and the raw material goes to zero. The engagement doesn't degrade evenly. It degrades on one side only, and it's the side that makes the content worth reading.
Every other client-side failure we've written about — approval committees, lane fatigue, the founder drifting up into abstraction — is a choice someone made, however unconsciously. This one isn't. The founder didn't decide anything. Their October just arrived.
Ecommerce founders have the most predictable collision calendar of any client type
This is the part that makes the problem solvable rather than tragic.
A SaaS founder's crunch weeks are semi-random — a funding round, a churn scare, an outage. An ecommerce founder's are on a schedule, and most of them are on the same schedule:
- Peak season prep, roughly August through October, when inventory is committed, deal submissions close, creative has to clear approval cycles, and the entire fourth quarter gets locked in.
- Peak season itself, late November through December, when nobody is thinking about anything.
- Post-peak reconciliation in January — returns, reimbursements, the annual fee changes, the accounting.
- Trade shows and buyer meetings, which cluster and which consume the week before and the week after, not just the days on the badge.
- One-off structural events — a 3PL migration, a rebrand, a raise, a channel launch, a supplier switch. These are unplanned in the calendar sense but entirely knowable once they start.
A founder doing $200K a month on Amazon can usually name eight to twelve weeks of the year when they become genuinely unreachable. We have never once had a client fail to name them when asked directly. We have also, historically, not asked early enough.
The tell that separates a collision from fatigue
These two problems produce similar-looking symptoms and require opposite responses, so getting the diagnosis right matters more than almost anything else in the relationship.
Lane fatigue produces argument. A bored founder edits. They sand the specificity off a draft, they push toward broader topics, they send you competitor content from bigger accounts outside the niche. They're engaged — they're just engaged in the wrong direction.
A bandwidth collision produces silence and speed. Drafts come back approved in minutes with no notes. Voice syncs get rescheduled and then shortened. Requests for a specific number get answered with "whatever you think." The founder isn't pushing back on anything because pushing back requires reading carefully, and reading carefully is the thing they don't have.
Here's the uncomfortable version: fast approval with no edits reads like trust, and it's usually the first symptom of absence. We've been fooled by it. A month of frictionless approvals feels like a relationship hitting its stride. It's often a founder skimming on a phone between calls.
The wrong prescription is the expensive part. Read a collision as fatigue and you respond by proposing fresh angles, new formats, a topic expansion — you hand more decisions to someone who has no capacity to make them. That accelerates the collapse. The founder now has an extra thing to think about in the month they had nothing left.
What a thin month actually costs
We'll keep this short because the mechanics of account-level reach recovery are their own subject.
Two costs stack. The first is distribution: a gap is an account-level withdrawal, and the posts after it pay for it before things normalise.
The second is the one founders feel and never trace. Content works on a lag. Buyers in this market lurk for months before they surface, and they surface in a DM with no post attached. So an October hole doesn't show up in October. It shows up as a quiet January — which is precisely the month a founder who just came off peak wants pipeline, has cash to deploy, and is asking why LinkedIn isn't working.
It is working. It's reporting on October.
Bank against it, the same way you'd bank inventory
Ecommerce founders already have the mental model for this. They just haven't applied it to content, because content doesn't feel like a thing that can stock out.
1. Build the collision calendar in week one. Not month four. The onboarding question is direct: which weeks of the next twelve months will you be unreachable, and what causes it? Put those weeks on a shared calendar. This takes eleven minutes and it is the single highest-leverage thing in the whole onboarding.
2. Bank in the quiet months, not the loud ones. The instinct is to try to write more when things get busy. That is exactly backwards — you're asking for extraction capacity at the moment it's lowest. The reserve gets built in May and June, at a normal cadence, from material the founder has anyway. Four surplus posts a month across two quiet quarters covers a peak-season hole comfortably.
3. Bank the durable stuff, not filler. A reserve full of generic "5 lessons" posts is worse than nothing — it publishes the founder's weakest content in the window where their profile is doing the most unattended work. The material that banks well is judgment and pattern content: what breaks in a category and why, what they believe now that they didn't two years ago, how a mechanism actually behaves. Tactical posts expire; judgment posts sit in a drawer for eight months and come out fine.
4. Lower the activation energy instead of the cadence. The 45-minute voice sync is the first casualty of a busy month, and cancelling it costs a month of raw material. The replacement isn't a shorter call — it's a different format. A six-minute voice memo recorded walking out of a warehouse, or three sentences dropped in Slack, produce usable material. We would rather have four 90-second voice notes in October than one immaculate call that gets moved twice and then dropped.
5. Pre-decide the floor. Cadence never degrades gracefully on its own; it collapses. Agree in advance what the reduced-cadence month looks like — two posts a week, drawn from the bank, no new extraction required — and write it down. A named floor gets held. An unnamed one becomes zero, and the founder feels like they failed, which makes them avoid the next call, which makes it worse.
6. Harvest the collision itself. This is the counterintuitive one and it's where the best material in the entire engagement lives. The weeks a founder is drowning are the weeks the most interesting things are happening — the vendor who missed a cutoff, the decision made at 11pm with incomplete information, the thing that broke during peak. Nobody captures it, because capturing it feels like a task. One 90-second voice note at the end of a hard day, no structure required, is worth more than a scheduled interview in a calm week. It doesn't get written that month. It gets written in February, when it's a receipt nobody else has.
The number to watch is weeks of supply
Stop measuring the reserve in "posts we have banked." Measure it the way the founder measures everything else: weeks of content on hand at the agreed cadence.
Under two weeks going into a known collision window is a stockout waiting to happen. Four to six weeks is comfortable. And when the number is falling in a quiet month, that's the signal to act — not when it hits zero in the middle of Q4.
The clients who come through peak season with their profile intact are not the disciplined ones. They're the ones who treated content like inventory and built safety stock in June.
FAQ
Won't banked content feel stale by the time it publishes? Only if you bank the wrong type. Anything tied to a current platform mechanic, a fee schedule or a live event has a short shelf life and shouldn't go in the reserve. Judgment, patterns, category behaviour and old-but-dated receipts hold for months. If a post needs a date on it to be true, date it and bank it anyway — that's more credible than an undated claim, not less.
Can't you just write without the founder for a month? We can produce something. It will be competent and it will be noticeably thinner, because the specificity is the product and the specificity only exists in their head. The honest framing is that a month without input costs you the thing you're paying for. The reserve exists so that month gets covered by their material, banked earlier, rather than by our best guess.
What about collisions we didn't forecast? The 3PL migration nobody saw coming, the tariff change, the acquisition conversation. Same mechanism, less warning. This is exactly what a standing reserve is for — it's not a Q4 tool, it's a buffer that absorbs whatever arrives. If you only bank against known windows you're still exposed to the unknown ones.
How much reserve is enough? Four to six weeks at your normal cadence, held permanently, topped back up after each drawdown. Beyond that you're pre-writing content that will age before it ships. Below two weeks you're one bad month from a gap.
Is this just a discipline problem? No, and framing it that way is why it keeps happening. A founder who runs out of hours during peak season is not undisciplined — they're doing their job. The failure is that nobody built the system to absorb a completely predictable event. That's a planning gap, and it's ours to close as much as theirs.
If your content has died in the same three months two years running, the fix isn't more discipline in October. It's a reserve built in June and a calendar somebody actually looked at. That's the kind of thing we set up in week one — get in touch if you want your Q4 handled before it arrives.