Who actually makes the reels, your social media manager or your editor?

Retention drops on a page running eight reels a month, and the meeting that follows has a shape you can predict.

The editor says the videos are fine and the posting times are wrong. The manager says the posting is fine and the hooks are weak.

Both are describing the same four decisions and neither of them is making any of them.

That gap is the whole reason VALORAE Media does the editing and runs the page as one function rather than two, and I want to lay out the actual split so you can check your own setup against it.

We learned it the way most people do. By owning one side of a handoff and watching a good video fail for a reason that was not on our side of the line.

Play 1: watch what happens when retention drops

The useful thing about the blame meeting is that it tells you exactly where the gap is.

Nobody argues about colour grade or about reply speed, because those have clear owners. The argument is always about the hook, the cover, the caption and whether to re-cut.

Four things. Always the same four.

Play 2: list the nine jobs inside one reel

Write them out and the ownership problem stops being abstract.

1. Decide the topic and where in the footage it lives.
2. Select the hook, meaning the first two seconds and the first line said.
3. Cut the body and pace it.
4. Caption it and check the burn against the safe zones.
5. Choose the cover frame.
6. Write the caption text and the on-screen hook card.
7. Choose the slot and publish it.
8. Reply to comments and DMs underneath it.
9. Read retention and decide what happens to the next one.

Jobs 3, 4 and part of 1 are unambiguously the editor. Jobs 7 and 8 are unambiguously the manager.

The rest is the gap.

Play 3: find the four with no owner

Hook selection. The editor picks a start point from the footage. The manager knows what the audience responded to last month. In a split arrangement the editor picks in the dark and the manager receives it too late to change.

Cover frame. Editors export a video and think of the cover as a thumbnail problem, which is a posting job. Managers get a file with no cover specified and pick whatever frame is sharpest. Nobody picks the frame that earns the tap.

Caption-to-edit match. The caption promises something and the edit delivers something. When two vendors write them, they drift, and a mismatch between promise and payoff is a three-second exit.

The re-cut decision. This is the expensive one. A video underperforms, and in a split arrangement the editor never learns, because retention data lives in the manager's login. So the same mistake ships again next week.

Retention is downstream of all four. Which means retention has no owner when the work is split, however good both vendors are individually.

Play 4: understand why the job descriptions are no help

Search the comparison and what ranks is hiring-site copy. Duties, required skills, average salary, day in the life.

That copy is written for the person taking the job, not the person buying the outcome. It is accurate and it is answering a different question.

A job description lists what a role does. It cannot list what falls between two roles, because from inside either role those four decisions look like someone else's.

Which is why the buyer reading those pages comes away thinking the choice is editor or manager, when the real choice is split or combined.

Play 5: publish the split in writing

Whatever you decide, write it down as a table your vendors both have.

For each of the nine jobs, one name. Not a team, a name. For the four contested ones, state who decides and who gets consulted.

Then add the piece that makes it real: the underperformance rule. Say that any video below a stated retention floor comes back to the person who cut it, with the retention graph attached, before a replacement gets made.

Contract clauses are worth getting right and this one belongs next to ownership and notice, because it is the clause that decides whether anything improves month to month.

Play 6: what changes when one team owns both

Three things change and none of them are glamorous.

The hook gets chosen by someone who saw last month's retention, so it is informed rather than guessed. The cover gets chosen against the caption rather than against the export. And a bad video comes back to the person who made it with the data attached, which is the only mechanism by which editing gets better at a specific account.

That last one is the real argument. A split arrangement has no feedback loop, so month six looks like month one with more files in it.

At VALORAE Media the editor sees the retention graph on their own video. That is the difference the org chart makes, and it is not a talent difference.

Play 7: decide whether to merge or to appoint

You do not have to combine vendors to fix this. You have to give the four decisions an owner with authority on both sides.

Merge if you are producing eight or more videos a month, if you have already had the blame meeting, or if you are the person currently arbitrating between two vendors every week.

Appoint if one vendor is genuinely excellent and you are not ready to move. Give one of them the four decisions in writing, give them read access to the other side's data, and accept that the other party is now taking direction.

What does not work is leaving it implicit and hoping. That is usually the story behind an agency relationship that quietly failed, told six months later as a personality problem when it was a structural one.

If you are still working out which single function you are short of, the editor or manager test is the earlier question, and the tier question is the one after it.

The short version

• Watch the blame meeting, because the four contested decisions are always the same four.
• Write the nine jobs down and put one name against each.
• Assign hook, cover, caption match and re-cut explicitly, since nobody claims them by default.
• Give whoever owns retention access to both the footage and the analytics.
• Add an underperformance rule that sends weak videos back to the editor with the graph attached.
• Merge the function at eight videos a month or once you are arbitrating weekly.
• Stop comparing the two invoices and start counting the decisions you are making yourself.

VALORAE Media edits the videos and runs the page as one function, which is the only arrangement where retention has an owner.