The category has a presentation problem. Agencies describe themselves as creative partners, platforms describe themselves as scale, and both descriptions skip the question a buyer actually has, which is what the money buys and where it goes.
Worth answering that plainly, including the parts where an agency earns its margin comfortably.
What an agency actually does
Five functions, bundled into one invoice. Separating them is the whole exercise, because you may need two of them and be paying for all five.
Sourcing
Finding creators who match the brief, in the right market, with the right style. Real work, and the function most easily bought elsewhere — creator platforms and marketplaces have made the supply side broadly accessible.
The brief
Turning a marketing objective into instructions somebody can film against. This is where the difference between usable and unusable footage is decided, and it is consistently undervalued by buyers comparing agencies on price.
Rights and clearances
Negotiating what you may do with the footage, for how long, in which territory, on which channels, plus music clearance and disclosure compliance. Unglamorous, occasionally decisive, and the function most often mishandled when buyers go direct.
Review and revision
Checking assets against the brief, sending them back, and absorbing the awkward conversations. Straightforward at ten assets, genuinely difficult at two hundred across eleven markets.
The fifth is creative direction — deciding what the campaign should look like at all. Some agencies are excellent at this and it is worth paying for. Many are reselling a template, and the way to tell is to ask what they would change about your brief before they have seen your budget.
Buyers compare agencies on the sourcing, which is the commoditised part, and pay for the brief without noticing that is what they came for.
The three routes, priced honestly
| Route | Typical cost per asset | Suits | Real constraint |
|---|---|---|---|
| Agency | $150 – $600 | Small volumes, negotiated rights, creative direction | Margin scales with volume even when the work does not |
| Platform or marketplace | $60 – $250 | Many assets, many markets, coverage over polish | You own the brief and the review |
| Direct outreach | $40 – $150 | Ongoing relationships, one or two creators | Rights and admin land on you, and they are the risky parts |
Those bands assume standard organic usage. Every one of them moves substantially once paid amplification enters the conversation, which is the subject of the next section and the reason quoted comparisons so often turn out not to be comparisons at all.
Direct outreach looks cheapest and frequently is not, because the two functions it strips out — rights and review — are precisely the ones whose absence is expensive rather than merely inconvenient. Somebody has to do them. If that somebody is a marketer at your company doing it for the first time, the saving is smaller than the invoice suggests.
The clause that decides the real cost
Usage rights are where UGC pricing actually lives, and a quote without them stated is not a quote.
Three tiers, and the gap between them is much larger than the gap between one creator and another.
Organic use only
You post the asset on your own channels. Cheapest, and adequate for a great deal of what brands actually need.
Paid amplification
You run the asset as an ad. Materially more expensive because the creator's likeness is now working in a media buy rather than a post, and rates commonly double or better.
Creator handle whitelisting
The ad runs from the creator's own account, borrowing their credibility. The highest tier, requires explicit agreement, and is the one most likely to be assumed rather than negotiated.
Two dimensions sit alongside the tier and both get skipped. Term — six months, twelve, perpetual — and territory. Perpetual worldwide paid rights on a face is a big ask and should be priced as one; a fixed term in three markets is a normal commercial conversation.
Get the tier, term and territory into the brief before anybody films. It costs nothing at that stage and cannot be retrofitted cheaply. What creators actually charge once rights are priced properly sets out the ranges from the other side of the table.
Where agencies are genuinely worth it
Being fair about this matters, because the honest answer is not "never".
A single hero asset with real creative ambition. Where one video carries a launch, the direction, casting and revision cycles are worth an agency's involvement and the margin is a small proportion of what is at stake.
Regulated categories. Anything where a spoken claim can create liability — supplements, financial products, medical devices, alcohol — benefits from somebody whose job is to catch a claim before it goes live. That is a specific competence and the cost of not having it is not measured in production fees.
Ongoing programmes with a consistent look. Maintaining a recognisable style across months of output is harder than it appears and is genuinely agency-shaped work.
And any situation where your team does not have time. That is a legitimate reason to buy a service and it does not need dressing up as anything else.
Where the model strains
Two situations, and both are volume and coverage problems rather than quality ones.
Many markets at once. An agency's roster is deep where it is deep. Asking for twenty creators across eleven countries, filming on the devices people in those countries actually own, tends to produce either a longer timeline or a quiet substitution of the markets they can service for the markets you asked for. How each layer of targeting narrows the pool and moves the rate sets out why that constraint is structural rather than a failure of effort.
High volume where coverage beats polish. Thirty adequate assets from thirty different people frequently outperform three excellent ones, because variety is what sustains a paid social account. Agency margin scales with the number of assets while the marginal work does not, so this is where the fee stops tracking the value.
Why "vetted creators" is a weaker signal than it sounds
Every supplier in this category says it, and it means less than buyers assume.
Vetting establishes that somebody has produced acceptable work before. It does not establish that they will produce what you need, because the variable that actually determines that is the brief. A strong creator working from a vague brief delivers footage nobody can cut. An ordinary creator working from a specific one delivers something usable most of the time.
What a brief needs is concrete and short: the hook, what must happen in the first three seconds, the shot list, orientation and aspect ratio, the claims that may and may not be spoken, the disclosure requirement, and what the asset is for. How to write one that produces usable footage covers the structure in full.
The second thing that predicts a usable asset is what the creator is asked to send back. Raw footage as well as the edit, in the specified resolution and orientation, is the difference between an asset you can recut for three placements and one you can post once. That belongs in the requirement rather than in a hopeful follow-up message. What counts as evidence and how to specify it before the work starts is the general version of the same discipline.
Vetting selects for people who have done good work. The brief decides whether they do good work for you.
What to ask before signing
Six questions, and the answers describe an agency more accurately than a deck of previous campaigns.
What exactly is the markup on creator payment, and is it disclosed? Some agencies are transparent about this and some are not, and it is a reasonable thing to ask directly.
Which usage tier is included, for what term and territory, and what does the next tier cost? A quote that does not answer this has not quoted your campaign.
Who writes the brief, and what would they change about ours? This separates creative partners from resellers faster than anything else.
How many creators can you actually field in each market we need, this quarter? Not roster size in aggregate — the number for your specific markets.
What comes back with each asset: the edit only, or raw footage, in what format?
And what happens when an asset does not meet the brief? Revisions included, timeline, and who absorbs the cost when the creator disagrees.
Rates, so the comparison is real
Per asset, standard organic usage, before markup.
| Deliverable | Creator payment | With paid-ad rights |
|---|---|---|
| Simple first impression or unboxing | $40 – $80 | $120 – $250 |
| Standard short-form clip to brief | $80 – $150 | $250 – $500 |
| Scripted demonstration, multiple takes | $150 – $300 | $400 – $750 |
| Photo set, 5–10 images | $80 – $250 | $200 – $500 |
| Monthly retainer, 8–12 assets | $1,200 – $3,500 | negotiated |
An agency markup sits on top of those, so the same simple unboxing lands somewhere around $110 to $200 once the margin is applied, against $40 to $80 plus a platform fee where you write the brief yourself. Whether that difference is worth paying is a real question with different answers for different campaigns, and anybody telling you it always is or never is has something to sell.
Choosing, in one paragraph
If you need direction, rights negotiation in a regulated category, or a single asset that has to be excellent, an agency is a sensible purchase and the margin is defensible. If you need volume, market coverage, or the same brief executed by many different people on their own devices, the margin is buying you something you are not using, and sourcing creators directly against a brief you control does the job for less. The comparison of the platforms in this space covers the middle route, and for the adjacent question of getting product into creators' hands without commissioning assets at all, how seeding actually works is a different mechanism with a different cost base.