Partway through a coaching call in August, I asked a client whether the first article I’d written for their blog had gone out that day.
They’d been out of the office the day before and hadn’t heard, so I pulled up their blog while we talked and went looking for it. It wasn’t there. The article was the first of a batch I’d written for them, the links were already in it, and the only job left on their side was adding a featured image. It was also sitting behind a calendar the team had filled so far ahead that, in their words, “we still have two weeks before September, but we already have things scheduled out and the next gap is in September.”
Nobody on that team was lazy, and nobody had rejected anything. The post was approved and finished, and it still hadn’t reached a single reader.
Approved content goes unpublished when a team’s shipping capacity, the number of pieces per week that can clear review and go live, is lower than what the team produces, and the gap almost always sits in the small jobs after approval that nobody owns.
Quick Take
- The stall happens after sign-off, in featured images, internal links, scheduling and the upload itself. Each of those takes a few minutes, and none of them has a name next to it.
- A full content calendar looks like health from the inside. A team that’s booked a month out has also made every new piece wait a month, including the best one.
- An approved post in a queue gives a buyer zero minutes with your thinking, so every week it sits is a week of marketing you paid for and didn’t get.
- The fix is to make approval the last human step. The approved file arrives ready to publish, one person owns the date, and the number you check each week is how many posts went live.
In this article
- Why approved content stalls
- What shipping capacity is
- How to tell your content is stuck
- Whether you should publish less
- How to raise shipping capacity without hiring
- What unpublished content costs an agency
Why does approved content never get published?
Because approval feels like the finish line, and in most agencies there’s still a handful of small jobs after it that don’t belong to anyone. Somebody has to make the featured image. Somebody has to add internal links, pick a date that doesn’t collide with something already queued, and push the thing into the CMS. None of those takes long, and a small task with no owner and no deadline loses to anything with a client attached to it.
That’s the same mechanism behind Delivery Drag, just further down the line. Client work has deadlines and your own marketing doesn’t, so marketing loses every calendar negotiation. With approved content, the writing already won its fight, and the finishing work loses the next one.
The client on that call had already found one of these and fixed it, and I loved that they had. “We simplified the heck out of the featured image, and just made it text for now,” they told me, “because we were spending way too much time creating custom images for each of the blogs, and it just wasn’t worth it.” They were right (most readers barely look at the image). But that’s one job of several, and my guess on the call was that the people doing the uploading couldn’t even see the calendar with the publish dates on it.
What is shipping capacity?
Shipping capacity is the number of pieces your team can take from approved to live in a normal week, with the finishing work done. It’s the only content number a buyer ever experiences, because a buyer can’t read your drafts folder.
I think of it as a sibling of Delivery Drag. Delivery Drag is marketing that never gets made because client work eats the time. A shipping capacity gap is marketing that got made and still didn’t reach anyone. The two feel completely different from the inside (one feels like being too busy, the other feels like being ahead), and they land in the same place: a pipeline that thins for reasons nobody can point to.
I’d bet the gap has gotten wider for most teams this year. Drafting got cheap and fast, and the steps after drafting mostly didn’t. That same client had connected their WordPress site to an AI tool so content could be pushed straight up, and when I asked about it, the answer was honest: “In theory. I don’t know if we have the team trained up and have the systems fully in place to simplify it that much.” Their production capacity went up, and their shipping capacity stayed right where it was.
And I’m not writing this from above the problem. My own content system has produced finished posts faster than I’ve approved and published them, and some of them sat for months before anything happened to them.
How do you know your content is stuck between approval and publishing?
The fastest tell is the question I asked on that call. If you can’t say whether today’s post went live without going to check, nobody on your team owns the publish step, and it’s running on whoever happens to remember.
A calendar booked weeks out is the second tell, and it’s the one that fools people, because it looks like discipline. The client described it perfectly: “We work ahead. To our detriment, sometimes. So then it’s hard to turn the cruise ship.” A deep queue means every new piece waits behind the old ones regardless of which is better, and the post you wrote with research behind it sits under a post somebody wrote to fill a Wednesday.
Then look at what did go out. If published posts are missing the finishing work, the team is shipping by cutting the steps that make a post useful. When I opened the post that went live the day before our call, it had no internal links at all. To Google, that reads as a standalone page with no reason for anyone to get to it, and pages like that often don’t get indexed.
And count the pile. If the number of approved-but-unpublished pieces is bigger this month than last month, you have a shipping capacity gap, however healthy the calendar looks.
Should you publish less content?
Sometimes, and the client asked the exact question I’d want every agency owner asking. They were already publishing three posts a week, I’d slotted mine for Tuesdays and Thursdays, and they worried about the math: “With implementation time, trying to do five, is that going to prevent us from getting the best two out there in a timely manner?”
My answer was to keep all five. “The only things that should have to happen to my articles is you guys add your image for the featured image,” I told them, because the links were in and the research was done. Adding two posts that cost the shipping step almost nothing is free volume.
Adding two posts that each need their own round of finishing work is a different decision. Those add pile, and the pile pushes your best work later. So the rule I’d use is simple enough to run on a Monday: add volume only when the new pieces arrive ready to publish, and if they don’t, publish fewer and publish the strongest ones first.
How do you raise shipping capacity without hiring?
Make approval the last human step, and then remove everything that currently happens after it. Hiring a person to push posts live mostly moves the pile to a new desk, for the same reason hiring a marketer before the system works usually disappoints.
Start by changing what “approved” means. A post isn’t ready for approval until the internal links are in, the meta description is written, the image is specified and the date is picked. When the approver says yes, the post should be able to go live in the next minute, and the approver’s job shrinks to the review pass at the end of the 10-80-10 rule. That’s how I delivered that client’s articles, and it’s why the only open job on mine was an image.
Then give the publish date an owner, by name. This is exactly the model in treating your agency like your own best client: somebody runs the account, and the founder approves at the altitude a client would, instead of sitting on every draft. If every post waits for the founder, you’ve rebuilt the owner bottleneck inside your marketing.
Cut the finishing work that doesn’t pay for itself. The text-only featured image is the model here. Ask of every step after approval whether a buyer would notice if it disappeared, and drop the ones they wouldn’t.
Automate the push where you can, and keep a human on the date. For another client on WordPress, I set up a connection that can only create drafts, so the upload happens automatically and a person still decides when it goes live. On my own blog, when a post goes live I ask my AI to find older related posts and add links pointing to the new one. As I told that client, if you’re doing the content anyway, spending an extra five to 10 minutes to get results from it is an easy trade.
Finally, put a different number in front of yourself. Each week, report how many posts went live, right next to how many were approved.
What does unpublished content cost an agency?
More than the hours that went into it. A buyer builds trust through time on brand, the minutes they spend with your material before they reach out, and a post in a queue contributes zero of those minutes. So the cost of a month-long queue is a month of buyer attention you already paid to earn, and it compounds the same way pausing marketing does. A post nobody can read also can’t be cited by ChatGPT or Perplexity, however good it is.
This isn’t a niche problem, either. In the Content Marketing Institute and MarketingProfs outlook for 2025, 33% of B2B marketers named managing workflow and the content approvals process as a challenge, down from 41% the year before, across 980 respondents (Content Marketing Institute). A third of the field is fighting the step between finished and live.
For agencies there’s a second cost, and my client named it before I did. Once we’d looked at how far back the missing links went on their own blog, the worry was whether the same gaps were showing up in the SEO work they run for clients. “That’s alarming to me,” they said. The agency’s own account is usually where shortcuts show up first, because it’s the one account with nobody on the other end complaining, and it’s the same reason content loses to client work in the first place.
For Monday, open your calendar and count two numbers for the last four weeks: pieces that got approved, and pieces that went live. If the first number is bigger, find the job that happens after approval with no name next to it, and put a name next to it this week.
And if your team is producing more than it can ship, I’d take that over an empty calendar any day of the week. It’s the easier problem to fix, and you’ve already done the hard part. If you’re sitting on a stack of approved posts and want help working out where they stall, bring it to the Dynamic Agency Community.
Where to go next
If your posts are going live on schedule and still nothing comes back, the bottleneck has moved upstream. Why your marketing isn’t working even though you’re doing all the things is about that case, where the activity is real but the positioning underneath it, or the room you’re publishing into, is off.
And if you’re deciding which pieces deserve the open slots in your calendar, another stall in the same market makes the case for the ones that explain your category to buyers. Buyers shop with the criteria of whoever explained the problem to them first, and AI answers are increasingly built from those same explanations.
FAQ
How many blog posts a week should an agency publish?
As many as your team can finish and push live every week without skipping the finishing work. For most small agencies that’s fewer than they can write. Pick the number you can hold for a quarter, then add only pieces that arrive ready to publish.
Who should own publishing on an agency’s own blog?
One person, by name, who isn’t the founder. The founder approves. Someone else owns whether it ships on the date, the same way an account manager owns delivery for a paying client.
Is it bad to schedule content months in advance?
Not by itself. A deep queue costs you flexibility, though. When a stronger piece shows up, it waits behind everything already booked, so keep open slots in the calendar for work that deserves to jump the line.
Can AI fix a content approval bottleneck?
Partly. It’s very good at the mechanical steps after approval, like pushing drafts into a CMS and finding internal links. It can’t give the publish date an owner, and a tool that’s connected “in theory” ships nothing until somebody’s job includes running it.
What should a post include before it goes to approval?
Everything it needs to go live: internal links, a meta description under 160 characters, the featured image or a spec for it, and a proposed publish date that doesn’t collide with the queue. If approval can’t be the last step, the post wasn’t ready for approval.