Client feedback is rarely criticism of you - it is unstructured requirements arriving through the only channel the client had. The craft of receiving feedback is mostly process: engineer the review so notes arrive usable, translate vague notes into questions before treating them as instructions, and protect the cut with versions so revisions never become relitigation.
Prepare the review, and the feedback improves
Most "bad" client feedback is a symptom of an unprepared review. Before sending the link:
- Label the stage. "Rough cut - structure feedback welcome" prevents a color-grading note on a temp grade. Clients calibrate to the stage you name.
- Ask for frame-level notes. "Pause on the exact frame and comment there" converts "the beginning feels off" into "the b-roll at 0:12 runs long". The first note costs a meeting; the second costs 60 seconds. The send-for-review workflow covers this in full.
- Name the decision maker. Unlimited feedback, one approval. Otherwise conflicting notes become your job to arbitrate.
Translate before you execute
Vague notes are compressed requirements. Decode them with questions - asked on the frame, in the thread, where context lives:
| The note | The question that unlocks it |
|---|---|
| "It feels slow" | "Which section - 0:12-0:40? Do you want fewer shots, tighter cuts, or shorter overall?" |
| "Make it pop" | "Do you mean the grade (brighter/contrast), the pacing (faster), or the music (bigger)?" |
| "Something is off in the intro" | "Is it the first logo, the VO tone, or the shot order? Tap the moment on the link." |
| "Can it be more premium?" | "Is there a reference film whose feel you want? Even one frame helps." |
One clarifying question in the thread costs a minute. Executing a guess costs a revision round.
Triage: must, should, could
Not every note has equal standing. Sort the thread before touching the timeline:
- Must - factual and contractual: wrong dates, legal flags, brand violations, missing deliverable. Fix unconditionally.
- Should - clear creative direction from the decision maker. Fix, and note what changed.
- Could - preferences, contradictions between reviewers, or things that harm the cut. Discuss first; some of these deserve a professional pushback ("that cut is what makes the intro land - want me to try a version without it so you can compare?"). Versions make that conversation cheap: ship both, let them compare.
The professional pushback formula
"I hear the concern about X. Here is why I cut it this way: [reason]. I have uploaded an alternative as v3 so you can compare directly - if v2 still reads better, we go with v2." You respected the note AND the craft, and the version history carries the decision.
Protect the cut with versions
Every revision round should be a version in the stack - not a new file name, not an overwrite:
- The stack is the receipt. "You asked for the faster cut on Tuesday, here is v3 with the change" ends debates that would otherwise run for weeks.
- Resolved notes stay visible. When a stakeholder re-raises a settled note weeks later, the thread (and its resolution) is one link away.
- Approval lands on a version. The recorded approval is what separates "done" from "they stopped replying".
The agency-side version of this discipline is in our client approval workflow case study; the tooling behind it is Frooty's review and approvals.
When feedback is genuinely hostile
Sometimes the tone is bad, not just the notes. Separate the signal from the delivery: respond to the substance on the frames, ignore the tone entirely in writing, and move genuinely unresolvable relationship problems to a call. The review tool helps here too - threads about frames are harder to escalate into arguments than email prose.
Frequently asked questions
How do I respond to feedback I disagree with?
Acknowledge it, state your reasoning once, and offer a comparison version so the client decides with their eyes. Disagreement settled by versions ends in an hour; disagreement settled by argument sours the project.
What if two stakeholders give contradicting notes?
Surface the contradiction in the thread, tag both, and escalate to the named decision maker. Do not average between them - conflicting creative direction is a client-side decision you facilitate, not make.
How many revision rounds are normal?
Two to three structured rounds is healthy for most projects. More than that usually means the feedback loop is vague (fix the process: frame-level notes, named approver) or scope is drifting (fix the contract: named rounds, scoped extras).
Should feedback happen on a call instead?
Calls are good for tone and negotiation, bad for precision. The best hybrid: collect frame-level notes in the tool, then discuss the two or three contested ones on a short call with the threads open in front of everyone.