The videos have been delivered for eleven days. They are good, nobody says otherwise, and they are still not online. Someone is waiting on feedback from someone who is waiting on feedback from someone.
It is the most invisible cost line of a UGC campaign, and the easiest to reduce. Not by speeding people up, which never works, but by defining who has the right to say no, and about what.
One approval round costs more than the fix it asks for
A remark such as "we could cut two seconds at the start" takes the editor ten minutes. The round that produced it cost a send, a wait, a viewing, a meeting and a chase, that is to say often three days.
That is the only arithmetic to remember, and it inverts the intuition. The aim is not better remarks, it is fewer rounds of them. A list of fifteen remarks in one round costs less than three remarks in three rounds.
There is an unpleasant consequence and it is better said out loud: asking someone to look once and say everything is more efficient than asking for a quick opinion. The quick opinion produces a remark, the remark produces a round, and the round produces a second viewing that produces another. That is why a circuit with no rule lengthens on its own, without anyone working badly.
What can be fixed, and what cannot
The distinction governs everything else, and it is purely technical.
Fixable in the edit: a cut, a subtitle, a shot order, a length, music, a call to action, framing within limits. Those are legitimate remarks at any time.
Not fixable: the location, the person, the light, what was said to camera, and the angle of the script. A remark about any of those is not a remark, it is a reshoot, and it should have landed at the brief.
| Type of remark | Who legitimately owns it | Fixable in the edit |
|---|---|---|
| Cut, length, pacing | the team distributing | yes |
| Subtitles, call to action | the team distributing | yes |
| Wording of a product benefit | product marketing | yes, if the footage allows |
| Claim, mandatory disclosure | compliance or legal | yes, via subtitle or voice |
| Choice of person, of location | nobody, after the fact | no, that is decided at the brief |
| Overall angle of the script | nobody, after the fact | no, that is decided at the brief |
By contrast with a circuit where anyone may say anything, this table gives each approver a scope. The question is no longer "what do you think" but "do you have a remark within your scope".
The brief removes half the rounds
Most unmanageable remarks are decisions that were not taken beforehand. If nobody settled the tone, someone will settle it at the viewing, and they will settle it against what already exists.
Three decisions taken at the brief remove the bulk: the tone, what may not be said, and who approves what. Writing that last line into the brief itself, with names, is the most profitable gesture in the whole process.
The list of what may not be said deserves a word, because it is almost always missing while it already exists somewhere in the company. Sales teams and customer service know perfectly well which wordings cause trouble, which comparisons attract complaints and which benefits cannot be promised. That list is collected in one conversation, annexed to the brief, and it removes the most expensive category of remark: the ones that arrive after the shoot about what was said to camera.
Mandatory disclosures follow the same principle: they belong in the brief and not in the feedback. A disclosure added in the edit always shows: it is placed after the fact, it is not in the frame, and it damages the video instead of accompanying it.
A circuit that holds in three roles
Past three approvers, the delay grows faster than the quality.
The one who distributes
They approve that the video does the expected job on the intended channel. They are the only one who decides to publish or not.
The one who knows the product
They approve what is said about the product, and nothing else. It is the most useful approver and the most often forgotten, which produces videos that are correct in form and wrong in substance.
The third role is the one who carries the risk. They step in only on claims and disclosures, and only when there are any. Involving them in every video as a matter of principle is the most frequent cause of a three week circuit.
What the outside can bring, and what it is not
France has a professional self regulatory framework for advertising, carried by the ARPP, along with advisory services for advertisers. It is a useful resource when a campaign touches a sensitive subject or a regulated sector.
It is neither a mandatory approval step for a UGC campaign nor an automatic gate. The confusion exists and it costs weeks: a brand that believes it must "have approved" every video externally waits for an answer nobody owes it. What is mandatory comes from the law and is checked internally; the rest is a choice.
The three circuits you meet, and what they cost
Not all circuits are equal, and it helps to recognise yours before trying to repair it.
The serial circuit
Each approver receives the video after the previous one. It is the slowest of all, because delays add up instead of overlapping, and it is the most common because it looks orderly.
It is fixed in one move: send to everyone at once, with a common date. The video stops passing from hand to hand and is watched by three people in the same window.
The funnel circuit
One person centralises, arbitrates contradictory remarks and passes on a single list. It is the most efficient when that person has the mandate to decide, and the worst when they do not: they become one more relay.
So the question to ask is simple and rarely asked: can that person say no to a remark, or only forward it.
The implicit circuit
Nobody is officially an approver, and yet nothing goes out without three people having given an opinion in a meeting. It is the most expensive, because it has neither date nor scope, and it is fixed by writing down what already exists in practice.
When an agency runs the campaign
When an agency produces for you, the circuit doubles without anyone saying so: it approves internally before showing you. The delay you observe therefore starts after a first cycle you never see.
The only thing to ask from the outset is whether what you are being sent is a first version or one already arbitrated. On a first version your remarks are useful; on an already arbitrated version they reopen a discussion that has already happened, and that is the surest way to add a week.
The two rules that remove the most delay
They are mundane and almost nobody writes them.
- One single round of grouped remarks, with a date. Past that date, anything not raised does not block publication.
- Silence means agreement, after an announced delay. It is the rule that makes the most difference, and the one that takes the most nerve to set.
The objection is always the same: if someone misses the deadline, a mistake will slip through. That is true, and it is a trade-off taken knowingly. The current cost is certain and counted in weeks on every campaign; the feared cost is rare and recoverable, since a video is taken down in a minute.
What to measure
One figure, noted across three campaigns: the number of days between delivery and going live. Not the number of remarks, not the number of rounds, the raw delay.
It is what tells you whether the circuit is improving, and the only one everyone understands without explanation. If it does not drop after the scopes are written, the problem is not the circuit, it is the brief.
Sources
Checked on 12 September 2026. This guide is not legal advice. Professional self regulatory rules do not replace the law and do not constitute an obligation of prior approval.
