The cut comes back. Six people watch it. Four of them send notes. Two of the notes contradict each other, one is about a font that hasn't been designed yet, and one just says "not feeling it".
The editor now has to guess which of six people to please. They will guess wrong, because they aren't in your meetings, and version three will be worse than version one.
This is almost never a talent problem. It's a process problem, and it has three rules.
Rule 1: Notes against the brief, not against your taste
The most useful sentence in a review is one that points at the document you both already agreed on.
- Actionable: "This doesn't hit the message in section four: someone watching this still thinks we're the same as the cheap version."
- Not actionable: "I don't love it."
The second isn't rude, it's just unusable. It costs the editor a day of guessing, and whatever they change might not be the thing you meant.
If you find yourself unable to trace a note back to the brief, that's information. Either the brief was incomplete (in which case say so plainly, and treat the change as new scope rather than as a correction) or the note is preference. Preference is legitimate. It just has to be labelled, because "this is my taste, and I'd like it anyway" is a completely different conversation from "this misses the objective".
Rule 2: The right notes at the right stage
Every note has a stage where it's cheap and a stage where it's expensive. Giving it early is free. Giving it late is a day's work. Giving it too early (on something not yet made) is wasted twice, because you'll have to give it again.
| Stage | Talk about | Don't talk about |
|---|---|---|
| Rough cut | Structure, order, length, what's missing, does it hold | Colour, typography, music, mix |
| Second cut | Pacing within sections, specific lines, emphasis | Whole-structure rethinks |
| Fine cut | Graphics, titles, transitions | "Should this be a different film?" |
| Pre-lock | Typos, legal, logo, final polish | Anything structural |
A rough cut is supposed to look unfinished. Scratch audio, no grade, temp music, placeholder graphics. Notes about colour on a rough cut are notes about work that hasn't happened.
And the reverse: a structural note at fine cut invalidates the graphics, the mix and the grade all at once. If you genuinely need one, say so and expect it to cost. That's fair. What isn't fair is delivering it as though it were a small note.
Rule 3: One voice
Consolidate. One document, one sender, contradictions already resolved.
This is the rule people resist most, because it means someone has to make a decision, and the whole appeal of everyone sending their own notes is that nobody has to. But an editor handed seven independent opinions has been given your internal disagreement to arbitrate, and they have no standing to do it.
How to run it:
- Everyone watches independently and sends notes to one person.
- That person merges them and kills the contradictions. Where two notes genuinely conflict, the named final approver decides. That's why the brief names one.
- One document goes to the editor.
- That document is complete. No "one more thing" the next morning.
That last point matters more than it sounds. Trickled notes are the single most demoralising thing you can do to someone in post, because every trickle re-opens work they'd already closed.
How to write a note that gets acted on correctly
Three parts, in this order.
Where. A timecode. "At 1:12" costs you nothing and saves the editor from scrubbing. It's covered properly in timecode feedback.
What you're observing. Not the fix. The symptom. "This section loses me" is more useful than "cut ten seconds here", because the editor knows fifty ways to fix attention and you know one.
Why it matters. Ties it to the brief or to the audience. "Because someone landing cold won't know what the product is yet."
Say what's working, and be specific about it
Not out of politeness. Because an editor with no positive signal has no idea what's load-bearing, and the next version may well "fix" the thing you liked.
"The cold open is exactly right, don't touch it" is one of the most useful sentences in a review. It removes a whole area from the search space.
How many rounds
Name a number in the brief. Two is normal for most branded work.
"Until you're happy" is not generosity. It's an unpriced risk, and it gets priced eventually, usually badly and in an argument. A defined number also changes behaviour on your side: people give better notes when they know there are two rounds rather than infinite ones.
If you need a third round, that's fine. Treat it as a change request with a cost, not as a continuation.
The failure modes, named
Notes by committee. Covered above. The most common and most damaging.
The silent stakeholder. Someone senior who didn't watch round one appears at round three with a structural objection. Prevent it by making the named approver's attendance at round one non-optional.
Comparison to a reference that wasn't in the brief. "Can it feel more like [X]?" at review stage is a new brief. If X mattered, it belonged in the moodboard.
Fixing it yourself in the note. Prescribing the edit rather than describing the problem. You hired someone for their judgement; notes that specify the solution use them as a pair of hands.
Approving in a meeting room. On a projector, with the lights up, at a size and in a context nothing will ever be watched in again. Watch it the way the audience will: on a phone, with the sound where they'll have it.
Where Acumin fits
Delivery Rooms are built around exactly this: versioned cuts, and comments anchored to a frame rather than floating in an email thread. Every note lands on a timecode, versions are kept side by side so "which cut are we talking about?" stops being a question, and the whole review has one home instead of living across three inboxes.
The consolidation rule still needs a human. No tool decides which of your two contradictory notes wins.
How to use this tomorrow
Before you send your next set of notes, do one pass over them and mark each one B (traceable to the brief) or P (preference).
Send both, but label the P ones. It's a small act of honesty that changes how the whole review is received, and it makes the ones that matter unmissable.
Related: Timecode feedback is the mechanics of the "where" half. How to hire a videographer covers agreeing the round count before any of this starts.