Approved With Comments Is Not a Lock
Sep 14, 2026
"Approved — with a few comments." That sentence just landed in the channel under the Round 2 Figma link. The brand manager added a heart. Account replied "great, moving to finals." Traffic asked for the file name. The designer opened the comments: new hero crop, softer claim, different end-card hierarchy, and a note that "we might want to revisit the concept for the cutdown." You are the creative project manager. Half the room thinks the round just closed. The other half just reopened the job inside the same sentence that claimed to approve it.
This is not a feedback-hygiene problem. People can write clearer notes forever and still mix close with reopen in one line. The failure is structural: approval language that says lock and change in the same breath, with no named version and no exit criteria for what those comments are allowed to touch. Until you separate those objects, "approved with comments" will keep masquerading as Round exit.
The Wrong Diagnosis Is Always Clarity
The room will tell you stakeholders should be clearer, designers should ask better questions, or you need a feedback template with required fields.
You paste a comment rubric. You ban "a few notes." You ask everyone to label Must / Nice / Question. You run a lunch-and-learn on giving feedback. Thursday still arrives with a soft yes that reopens claim and layout, because clarity was never the constraint.
Clear comments on an unlocked object are still a reopen. Vague comments on a locked object are still a reopen if you let them move concept. The problem is not adjective quality. The problem is that the approval sentence carried two jobs: terminate the round, and keep changing the work. Creative teams treat that mash-up as progress because it feels collaborative. Traffic cannot traffic a mash-up. Legal cannot clear a mash-up. A freeze date cannot feed on a mash-up.
More templates will not fix this. A nicer tone will not fix this. This post is what to do when the channel says approved and the comments still reopen concept, claim, or layout.
Approval Language That Mixes Close and Reopen Is Not a Lock
That is the structural problem.
Most creative teams treat "approved" as a vibe word. It can mean "I like the direction," "ship it," "I am done looking," or "please keep going while I stay in the file." Then someone adds "with a few comments" and the room hears exit. Production starts finals. Status turns green. The comments, meanwhile, change the objects that made the round a round: concept direction, the claim legal already reviewed, the layout hierarchy the creative director locked last pass.
A lock needs three things in writing:
- A named version (file name or frame name, not "the latest")
- A named approver yes that ends the decision for that version
- Exit criteria for residual comments: what they may still change, and what they may not
Without those, "Approved with comments" is a polite reopen. It is the cousin of Status Green with concept still open, and the cousin of counting two rounds without a decision plan. The board can look healthy. The date can look healthy. The decision is not closed.
This sits next to five posts you may already know. Your Status Is Green. Concept Is Still Open. is when the board reports tasks while locks stay open. "Two Rounds" Is Not a Decision Plan is why counting passes is not exit criteria. The Kickoff Happened. Week Two Still Feels Like Day One. is how you leave locks behind at the start. The Media Buy Is Booked. The Creative Still Isn't Approved. is the hard-date collision. What to Say When a Client Asks for One More Thing is Friday scope arriving as collaboration. This post is the sentence that fools everyone mid-round: the soft yes that reopens the job while claiming to close it.
Split the Sentence Before Anyone Moves to Finals
Do not ban comments. Ban the mash-up. When someone writes "approved with comments," force the channel to answer four objects before production treats the round as closed.
1. Named version
Which file, frame set, or export is being decided. "This" and "the latest" are not names. If the brand manager approved v7 in Slack while the designer is still in a branch called v7-alt-crop, you do not have a lock. You have a misunderstanding with a heart emoji.
2. Close vs residual
Is this a close on concept, claim, layout, and copy for this version — with residual notes limited to production polish — or is this still an open round that happens to include the word approved? Make the person say which. Soft language loves to hide inside praise.
3. Allowed change list
Write what residual comments may still touch. Specs, crop for a required size, typo, legal line that does not change the claim, export naming. Then write what they may not touch without a named reopen: concept, primary claim, hero hierarchy, end-card story, audience, offer.
If a comment on that list would change concept, claim, or layout hierarchy, it is not a residual note. It is a reopen request wearing approval clothing.
4. Cost of reopen
If comments cross the allowed list, name the cost before design spends another hour: new round number, date slip, dropped size, overtime, freeze risk. You are not punishing feedback. You are refusing to let "a couple notes" invent a free second concept pass.
Put those four answers in the same thread as the approval sentence. A private DM to account is not a lock. Traffic and the designer need to see the same close.
What "Approved With Comments" Is Allowed to Mean
Use this definition in kickoffs and round briefs so the phrase stops being improvisation:
Approved with residual production notes means: named version is locked for concept, claim, and layout hierarchy. Comments may only change production items on an explicit allowed list. Designer may proceed to finals. Traffic may prepare specs. Status may say Round Closed / In Finals.
Approved in spirit, still open means: stakeholder likes direction but has not closed the decision objects. Round is not closed. Do not move to finals. Do not empty the Finals folder with hope. Rename the status to Round Open — Comments Pending Classification.
Anything between those two is fog. Fog is how Thursday happens.
Legal has a special case. If legal writes "approved with comments" and the comments change the claim, that is not residual. That is claim reopen. Visual polish on a claim that can still die is waste. Route claim comments through the claim gate, not through "design notes."
The Sentences That Separate Close From Reopen
A definition nobody defends is decoration. These are the lines to send in the channel when approval language tries to do two jobs at once.
1. The soft yes just landed
They say: "Approved — with a few comments. Looks great overall."
Do not send: "Awesome, moving to finals."
Send:
"Need to split this before we treat Round 2 as closed. Confirming: is [VersionName] locked for concept, claim, and layout hierarchy, with residual notes limited to [allowed list]? Or are the comments allowed to change concept/claim/layout? If they can change those, this is still an open round, not an approval. Reply with lock or reopen so traffic and design share one truth."
Why it works: you did not reject the stakeholder. You forced the sentence to pick a job.
2. Account translates praise into exit
They say: "Brand loves it. Just a couple notes. We are good to finals."
Do not send: "Perfect, I'll tell traffic."
Send:
"Love is not a lock. I need a named version, a written close from [approver] on concept/claim/layout, and a list of what the notes may still change. Until those three exist in this thread, Finals stays empty. I will not traffic a soft yes."
Why it works: account is often optimizing for relationship heat. You protected the relationship and the freeze path at the same time.
3. Comments reopen concept inside an "approval"
They say: "Approved with comments — also, can we explore a different hook for the cutdown?"
Do not send: "Noted, designer will take a look."
Send:
"That comment reopens concept for the cutdown. It is not residual polish on [VersionName]. Options: (A) lock the hero as approved and park cutdown concept as a separate scoped round with a date, or (B) keep Round 2 open and slip finals. Which does [approver] want in writing?"
Why it works: you named the reopen before design burns a day inside a false close.
4. Designer asks if they should just "knock out the notes"
They say: "I can hit these tonight so we stay on schedule."
Do not send: "Yes please, thank you."
Send:
"Pause. Two of these notes change hierarchy and claim. That is a reopen, not overnight polish. I need [approver] to confirm lock vs reopen before you spend the night. If we treat reopen as polish, we own the surprise when legal or brand rejects the 'approved' file tomorrow."
Why it works: you protected the designer from unpaid concept work dressed as helpfulness.
5. Legal comments land under an approval stamp
They say: "Legal: Approved with comments — soften the claim and add the longer disclaimer."
Do not send: "Designer will update and we'll call it approved."
Send:
"Claim change means claim is reopened, not residual. Visuals stay frozen against the old claim until the new line is written and cleared. I will not mark creative approved while claim is moving. [Approver] + legal: confirm the new claim text and return date, then design updates once."
Why it works: you kept claim and visual on separate gates so "approved" cannot launder a rewrite.
6. Someone wants the round number to advance anyway
They say: "Call it Round 3 start. The comments are small."
Do not send: "Sure, updating the board to Round 3."
Send:
"Round numbers do not create exit criteria. If Round 2 never closed — no named version lock, no allowed-change list — Round 3 is just Round 2 with a higher number. I will update the board when we have a written close or a named reopen with cost. Counting passes is not a decision plan."
Why it works: you tied this failure to the round-counting trap without turning the thread into a lecture.
7. Leadership asks why finals slipped after "approval"
They say: "I thought this was approved yesterday. Why are we late?"
Do not send, in the leadership thread: "Stakeholders keep changing their minds."
Send, privately to account or your lead first, then one clean line upstairs:
"Yesterday's note was 'approved with comments' that reopened claim and layout. We did not have a named-version lock or an allowed-change list, so production correctly did not treat it as close. We now have [lock / reopen decision] in writing. Finals date is [date] if the lock holds."
Why it works: you refused the blame story and reported the missing decision objects.
What Changes When You Refuse the Mash-Up
Rounds get shorter because residual notes stop pretending to be concept passes. Designers stop doing night work on false closes. Account stops selling "they're happy" as a freeze. Traffic stops speccing files that can still mutate. Legal comments route to claim gates instead of disappearing into design polish. Status can say Round Closed without lying.
In-house, there may be no client. There is still a brand manager, a creative director, and a channel full of soft yeses. Same split. If the director writes "approved with a couple notes" and one note changes the hero story, that is a reopen. Say so in the shared sheet before the launch date absorbs the fiction.
In production environments, lock language is already native. Steal it. "Is this locked, or are these comments a reopen?" belongs in Slack the same way it belongs in editorial. If the answer is reopen, do not advance the round number to soothe the room.
None of this requires a new tool. It requires refusing to let one sentence do the job of close and the job of change at the same time.
Frequently Asked Questions
What does "approved with comments" mean in creative project management?
On its own, it means almost nothing useful. It can be a real close with residual production notes, or an open round that includes the word approved. Force a named version, a written close on concept/claim/layout, and an allowed-change list. Without those, treat it as Round Open until classified.
How do you close a creative round when stakeholders leave comments?
Separate residual from reopen. Residual comments stay inside an explicit allowed list and do not touch concept, primary claim, or layout hierarchy. Reopen comments require a named cost and keep the round open. Do not move to finals until the classification is in the same thread as the approval language.
Who owns classifying approval comments — the CPM, account, or creative director?
The CPM owns the classification sentence and the board language. The named approver owns whether the version is actually locked. Account owns the client-facing conversation when a soft yes has to become a hard choice. The creative director owns whether a visual note is hierarchy change or polish. Align privately if needed, then put one classification in the channel.
Is a feedback template enough to fix soft approvals?
No. Templates improve comment clarity. They do not create exit criteria. A perfectly labeled Must-fix that reopens concept is still a reopen. Fix the decision objects first: named version, close vs residual, allowed list, reopen cost.
What should traffic do when creative says approved but comments are open?
Do not start specs off an unclassified soft yes. Wait for a named-version lock and an allowed-change list that excludes concept/claim/layout hierarchy, or wait for an explicit reopen decision with a new date. Empty Finals folders on purpose beat rework on ghosts.
Do I need a certification to run approval language this way?
No. You need the habit of splitting close from reopen in the same thread where the soft yes landed. A creative-specific credential helps if you want brief discipline, single-approver routing, and decision closure taught as one system. It is not required to refuse "approved with comments" as a fake lock.
Closing
If your rounds keep "closing" on sentences that still reopen concept, claim, or layout, the problem is probably not that stakeholders write messy notes. It is that approval language was allowed to mix terminate and change in one breath, with no named version and no exit criteria.
You do not need another feedback workshop. You need locks that can survive the word approved: named version, written close, allowed residual list, and a named cost when comments cross that list.
The default path is the Bundle: Level I ($147), Level II ($197), the Creative PM AI Kit ($97), and the Resume Kit ($57). $297 one-time. Bought separately those are $498, so you save $201. Self-paced. Unlimited exam retakes. Digital certificates emailed when you finish. They never expire. 5-day full refund if it is not the right fit. Enroll in the Bundle.
If you only want the foundation, Level I is the alternative at $147. Eight self-paced modules, typically 10 to 15 hours, 70% to pass, unlimited retakes. The credential is a Certificate of Completion, emailed when the requirements are met. It never expires. Start with Level I.
Not ready yet? Start with the free eBook, The Essentials of Creative Project Management.
Who should skip this: if your team already refuses soft yeses without a named version and an allowed-change list, and "approved with comments" cannot advance finals until classified, you do not need this post. Keep splitting the sentence. Come back when the mash-up returns.
The Only Certification Built for Creative Project Managers
CPMA is the only certification built exclusively for creative project managers, developed with veterans from Disney, Google, Snap Inc., Red Bull, Sony Pictures, Accenture, and Paramount Pictures. The Bundle is $297. Level I is $147 if you want the foundation only. Or start with the free eBook.