Video marketing agency cost - and when not to hire one

What a video marketing agency actually charges

The honest way to price a video is line by line, because that is how the work is actually sold. A rate card published by Harvard Medical School's IT department prices it that way. It is one institution's card rather than an industry average, but the line items are the ones every production quote is built from. One operator and a 4K camera: $750 for a half shoot day, $1,200 for a full day. Two operators: $1,350 or $2,200. A producer at $85 an hour, an editor at $110 an hour. A half day is five hours or less; a full day is nine or less.

Two things about those numbers decide what a video marketing agency really costs you.

First, the per-video number is not the number that matters to you. A one-off brand film is a single purchase. When you run short-form ads you are not buying one video. You are buying a supply of them, and the price of the next one is what governs your budget.

Second, there is a cheaper supply line for the part that is only a camera. If the whole job were "someone has to hold a phone", that is the cheapest line on the card. Which suggests the bottleneck sits somewhere else.

The cost that decides your budget is the revision

In production work, changes beyond the agreed scope are billed separately. On a rate card a revision is not a product, it is an editor's hours. That is $110 an hour, the same line item whether the note is small or large. Practitioners consistently name revision rounds and mid-project scope changes as the leading cause of budget overruns. Not the shoot, not the cast, not the edit.

You cannot know in advance which hook, which opening, which framing will land. Finding out takes attempts. The working estimate in media buying is one profitable creative per ten to twenty tested. So the activity you need most is precisely the activity an agency has to bill you for: changing one thing and running it again.

That is the structural problem, and it is not about price level. A cheaper agency has the same structure. So does a faster one.

The unit is what differs. If you buy deliverables, every test is a purchase. If you buy changes, testing is the cheap part.

How long it takes

The same logic applies to time. A production day is five hours or less for a half day, nine hours or less for a full day. Everything else is prep and post around it. Calendar time is dominated by how many of those days the work needs, and by how many rounds of notes come back between them. That is where the difference between video marketing companies shows up first.

The brief and the planning come before anyone stands on set, and rushing them is the biggest cause of rework later. Production itself, the part everyone pictures, is usually one or two shoot days. Everything after it is editing, at an hourly rate, one round at a time.

None of that is waste. It is what a considered, crewed, insured production costs in calendar time. It only becomes a problem when the thing you need is twelve attempts this month, not one considered film this quarter. A two-month cycle gives you six attempts a year, and six attempts is below the number the odds above suggest you need.

Does a retainer fix it?

The obvious next move is a retainer, and it is a genuinely better deal per video. It bundles shoot days and editing into a monthly figure and prices them below a one-off project. Run the arithmetic and it looks like a large discount, five or six videos for the price of one project video. If your need is a steady handful of videos a month, take it.

But look at what the retainer actually changed. It converted a price per video into a cap on volume: five or six deliverables a month. Everything above that is a scope conversation, and scope conversations in production are exactly where the revision charges live.

So the retainer answers "how do I pay less per video". It does not answer "how do I run twenty attempts this month to find out which hook works". Twenty attempts is four months of a standard retainer, and the month you ask for them is the month you renegotiate.

That is the same structural limit, just packaged monthly. A retainer is a better price. It is not a different unit.

Three options, not two

Most of this decision gets framed as "agency or in-house", and that framing is incomplete. Enterprise marketing teams confronting the same capacity gap describe it as three options: do nothing, hire someone, or find another way.

That third option is the one people skip, and "another way" is not automatically better. It depends on what you are actually short of:

Which one you are short of decides the answer. Most teams assume it is the first and discover later it was the third. It is also the question a video advertising agency rarely asks out loud.

• Short of people who can hold a camera? That is the cheapest line on the rate card, and the creator market has already made it a commodity.
• Short of someone to own strategy, brand and channel? Hire the agency. That work does not disappear just because production got cheaper.
• Short of enough attempts to find what works? This is the gap that neither hiring nor doing nothing closes, because both are priced per deliverable and you need changes, not deliverables.

Five questions to ask before you approve a retainer

No agency page will tell you this, for obvious reasons. But you are about to commit to a recurring spend, and the questions that change the number are not the ones about creative quality.

1. What counts as included scope, and what starts a new one? A retainer buys a fixed amount of work. Ask what the boundary is in writing: how many assets, how many revision rounds, how many channels. Everything on the far side of that line is quoted separately.

2. What happens when the brief changes in month two? Scope changes are the leading cause of budget overruns in this industry, ahead of production itself. Ask what the change process is before you need it, not after.

3. How many revision rounds per asset, and what does round three cost? This is the line that decides whether you are buying a service or renting a fixed number of changes. Get the number in writing.

4. Who owns what is made, and what can I reuse if we part ways? Ask whether you keep the scripts, the structures, the footage and the accounts. Or whether you leave with finished files and start from zero somewhere else.

5. What are you actually responsible for, output or outcome? A creative partner owns the argument; a media partner owns the spend; an execution partner owns delivery. Agencies are honest about which one they are, but the retainer document often is not.

The first three decide your budget. The last two decide what you own in a year.

What you are actually buying in an agency retainer

The retainer line items are rarely broken out, but a marketing engagement is usually three different things bundled together. Knowing which one you need changes what you should be paying for.

Judgement. Deciding what to say, to whom, on which channel, in what order. This is the part that does not get cheaper when production gets cheaper, and it is the part most worth paying for. If this is your gap, no tool closes it. You will just produce more videos that miss.

Execution. Making the assets. Briefs turned into finished work on a schedule. This is the part AI production has genuinely repriced, and the part where a retainer priced by output gets expensive fastest.

Placement and spend. Running the media, managing the accounts, optimising the buy. This is measured in budget managed rather than in videos made, and it is a different contract shape entirely.

Most agencies are strongest at one of the three and honest about the other two. The trouble starts when a retainer is priced as if all three were equally weighted, because then you are paying judgement rates for execution volume. That is the single most expensive way to buy attempts.

The difference is the billing unit

We are deliberately not putting our price next to theirs on this page. Prices move, and the number is not the point. The unit is. A per-deliverable vendor gets more expensive the more you test. A per-change vendor does not. That distinction is what separates a social media video agency pricing by output from one pricing by change.

The hundredth variant does not cost what the first one did, because a variant is mostly a re-render. A run generates up to ten variants at a time, and you can run it again for the next ten. The marginal cost does not climb.

"But our agency already uses AI"

Fair point, and worth taking seriously. Agencies are openly reworking their pricing around AI production right now. So "we use AI and they don't" would be a bad argument, and it is not the one here.

Using AI changes an agency's cost to produce. It does not change their billing unit, which stays attached to the deliverable. That is what the engagement is structured around: scope, rounds, approvals, invoices. An agency that produces ten times faster still sells you a video, and still bills a revision round when the brief moves.

That is not a criticism. It is what a services business is. It just means the AI part of the story is genuinely good news for their margin and only incidentally good news for your testing budget. That holds unless and until the engagement itself is repriced per change rather than per output.

So the test is simple, and you can run it on any agency: what do they charge for the fourth revision? If the answer is a number, the structure has not changed.

What stays with you

Three things move in-house when the unit changes, and they are the reason this is not just a cheaper vendor.

The starting point. You do not have to invent the format. Paste a link to a video that is already working in your market, a page URL is fine, no file needed. BrandU rebuilds its structure with your product, your claims and your voice. You keep the order of the argument; you replace everything that made it theirs.

The asset, not just the output. A commissioned video is a finished thing. A rebuild leaves you something you can keep editing and reusing, so next month starts from a working structure instead of a blank page.

The changes. You work on the script and the scene list rather than a timeline: no shoot, no editing timeline, no render queue. And because the video is built as independent segments, rewriting one line re-runs that segment and its voiceover while the rest is kept.

The eight things a rebuild separates, if you want the detail:

When you should still hire an agency

This is the part that matters most, so it is not a disclaimer.

Hire an agency when the work is the point. A brand film, a launch, a founder story, a piece that has to be right once. That is craft work, and craft work is what a good agency sells. Our whole argument is about volume and iteration. It says nothing about a film that only gets made once.

Hire an agency when nobody owns the strategy. Deciding what to say, to whom, on which channel, in what sequence. That is real work, and buying a tool does not do it for you. If that is genuinely your gap, a tool will not close it; you will just produce more videos that miss.

Hire an agency when you need a crew. Your video may have to show a real factory, real staff, a real kitchen, a product being made. Then someone has to be there with a camera. We do not shoot anything.

Stay with an agency when the relationship is working. If your agency is already delivering at the volume you need at a price you can defend, the honest advice is to keep them. Switching costs more than this page is worth to you.

What we are claiming is narrower than "agencies are obsolete". It is this. For the specific job of finding out which of ten versions works, the per-deliverable model is the wrong instrument. Using it that way is expensive.

What BrandU does not do

• It does not plan your marketing. No strategy, no channel mix, no media buying. If that is your gap, hire for it.
• It does not shoot anything. No crew, no studio, no location. If your video needs filming, we are not the answer.
• It does not guarantee the video performs. Nothing does. That is why the revision section spells out the odds.
• It does not make one-off brand films. It is a production line for iterations, not a replacement for a piece of craft.

How to test the claim cheaply

• Take one video that is already working in your market, not the most famous one, one that is working now for an audience like yours.
• Paste the link and rebuild it with your product and voice.
• Change one thing and run it again. Note what you changed.
• Compare the cost of that second attempt to the cost of a revision round. That comparison is the entire argument on this page, and you can run it before committing to anything.