What should a video brief include?

The five decisions above, plus the practical constraints: budget range, deadline, who approves, where the asset will live, and what already exists that it must sit alongside. A useful test is whether a stranger could read the brief and say what would count as success. If success is only describable as “a great film”, the brief is a mood board with a budget attached.

How long should a video brief be?

One to two pages. Length is not the quality signal — decisiveness is. A twelve-page brief that never names a funnel stage gives a production company less to work with than a single page that does. The 60-Second Brief is named for the test: if the five decisions cannot be stated aloud in sixty seconds, they have not been made yet.

Who should write the video brief, marketing or the agency?

Marketing writes the five decisions; the agency writes the treatment. That division is the point. The decisions are about your funnel, your stack and your pipeline, which the agency cannot see. The treatment is about story, craft and execution, which is what you are paying them for. Briefs go wrong most often when the client writes the treatment and leaves the decisions blank.

What is the most common mistake in a video brief?

Specifying the deliverable instead of the job. “We need a two-minute brand film” is a purchase order, not a brief — it names a format and a runtime and leaves the funnel stage, the viewer action and the success measure undefined. The asset then gets judged on whether people liked it, because nobody wrote down anything else it could be judged on.

Should the brief specify a budget?

Yes, as a range. Withholding it in the hope of a cheaper quote produces proposals scoped against a guess, which wastes both sides’ time and makes the responses incomparable. A range lets a production company tell you honestly what is achievable inside it, and lets you check the number against the acquisition cost the asset is meant to offset before anyone starts.

The 60-Second Brief

How to write a video brief for an agency

Five decisions before anyone picks up a camera.

By Paul Joseph · Updated 23 Aug 2026

A video brief for an agency needs five decisions on paper before spend: the funnel stage the asset serves, the single action you want a viewer to take, the martech trigger that fires behind that action, the contact record the view should touch, and the pipeline outcome the asset will be judged on. Everything else — tone, references, runtime, delivery formats — is production detail a competent agency can propose. Those five are the ones only you can answer, and if you do not answer them the agency will substitute its own, which is how a film ends up excellent and unaccountable at the same time.

If you are writing a brief because the last asset underperformed, start with the Dark-Funnel Video Audit instead — two minutes, free, and it tells you whether the problem was the brief or the plumbing behind it.

Why most briefs fail before they reach an agency

The typical corporate video brief opens with a company description, moves to an audience paragraph, lists three adjectives for tone, links a reference film, and names a runtime and a deadline. It reads as thorough. It contains no decision anyone can be held to.

That is not carelessness. It happens because the brief is usually written by the person who will manage the production, and the questions that actually determine whether the asset works — where it sits in the funnel, what the stack does when someone watches it — belong to other people. Asking them takes a week, so the brief goes out without them and the gaps get filled by whoever is closest to the work.

The cost of that is not paid at the brief stage. It is paid at the review, when a film that met every stated requirement is judged against a requirement nobody stated, and again a quarter later when somebody asks what it produced. A well-made film asked to deliver a result it was never designed to reach is the single most expensive failure in corporate video, and it is entirely preventable in an hour.

The five decisions

This is the Frame to Funnel Method's 60-Second Brief — named for the test rather than the length. If you cannot say all five aloud in sixty seconds, the asset is not ready to commission, whatever else the document contains.

01 · The funnel stage it serves

Which stage of the buyer's journey this asset is for, stated as one stage rather than a list. A film aimed at everyone is aimed at no step, cannot be placed anywhere specific, and will be reviewed by whoever happens to be in the room. Name the stage and the review question changes from “do we like it?” to “does it move someone from here to the next step?”
02 · The single viewer action

One thing you want the viewer to do next, and only one. Three calls to action produce none. The action has to be proportionate to the stage — asking for a demo at the top of the funnel fails not because the film was weak but because the request was mistimed. “Visit the website” is not an action; it is the absence of one.
03 · The martech trigger behind it

What your stack does the moment someone takes that action — the automation that fires, the list they join, the alert a seller receives. This is the decision most briefs never contain, and the one that determines whether the asset produces anything traceable. It also has a hard dependency: someone who administers the stack has to confirm it is possible before the brief is signed.
04 · The contact record it should touch

Where the view is written down, and against what. Not “web”, not “video” — the specific asset, on a specific record, in a field somebody can report on. If nobody can name the field, the view will be recorded nowhere and the asset will be undefendable at the next budget review regardless of how well it performed.
05 · The pipeline outcome it is judged on

The number this asset is accountable for, agreed with whoever owns that number, before production starts. It does not have to be revenue — demo requests from a named step, meetings booked, cycle time on deals the film touched. It has to be something that exists in a system, and it has to be agreed in advance, because a success measure chosen after delivery is a negotiation, not a measurement.

Read together, the five describe a chain: a stage implies an action, an action needs a trigger, a trigger writes to a record, and a record makes the outcome measurable. Break any link and the ones after it stop working, which is why answering four of five is not four-fifths as good.

A worked example: the same brief, twice

A B2B software company briefs a customer story. Here is what they sent first, and what they sent after the five decisions were made. The budget, the agency and the customer were identical in both versions.

The same video brief before and after the five decisions
Decision	First brief	After
Funnel stage	Not stated — “brand awareness and lead gen”	Mid-funnel: prospects who have had one call and are building an internal case
Viewer action	“Learn more about us”	Book a technical deep-dive from the end frame and the page beneath it
Martech trigger	None specified	Booking fires a Slack alert to the deal owner and tags the account
Contact record	None specified	Asset name written to a multi-value content-influence field on the account
Judged on	“Engagement and brand lift”	Deep-dives booked from the asset page, and cycle time on deals it touched
Runtime the agency proposed	4:30, because nothing constrained it	2:10, because a mid-funnel viewer with one call behind them needs less setup

The second brief is not longer. It is a page shorter, because naming the stage removed four paragraphs of audience description that were only there to cover the fact that the audience had not been chosen.

Note what changed on the agency's side without anyone asking. The runtime halved, because a defined stage tells a director how much setup the viewer already has. The end frame acquired a specific job. And the review, when it came, took forty minutes instead of three rounds, because there was an agreed question to review against.

The arithmetic underneath this matters too: at $12,000, the customer story has to offset roughly that much acquisition cost over its life to be worth commissioning. The second brief makes that checkable in advance — run it in the CAC-to-Frame calculator before signing. The first brief makes it unanswerable in principle.

What belongs in the brief, and what does not

Beyond the five decisions, a brief needs its constraints and nothing else: a budget range, a deadline, who approves and in how many rounds, where the asset will live, and what it must sit alongside. Five lines. Anything more usually crowds out the decisions rather than supporting them.

Three things that do not belong. Shot lists and treatments — you are buying the agency's judgement, and specifying execution while leaving strategy blank inverts the value of the engagement. A wishlist of deliverables — nine cut-downs decided before the master exists is a budget allocated to a guess. And reference films with no note on why — a link to a film you admire tells a director nothing unless you say which part you mean, and it frequently transmits an expectation of a budget you did not set.

Withholding the budget deserves its own line, because it is the most common false economy in commissioning. A production company scoping against a guess will either propose something you cannot afford or something smaller than you needed, and the responses become impossible to compare. Give a range. It costs nothing and it is the single change that most improves the quality of what comes back. The cost drivers are set out here if you need to sanity-check the range first.

How to get the five decisions when they are not yours to make

Decisions three and four usually sit with marketing operations, and decision five with whoever owns the pipeline number. That is the real obstacle: the brief is due Friday and those people have their own quarter. Two things make it tractable.

Ask for a decision rather than a discussion. “Can a view on this page write the asset name to the account record — yes or no, and if no, what is the nearest thing that works?” takes an operations person four minutes. “Can we chat about video attribution?” takes a meeting nobody schedules. And bring the outcome question to the pipeline owner as a proposal, not a blank: “we intend to judge this on deep-dives booked from the asset page — is that the number you would defend?” People correct a proposal far more readily than they author one.

If a decision genuinely cannot be made in time, write down that it was not made, and what was assumed instead. A brief that says “no trigger available this quarter; the asset is judged on page conversions only” is honest and reviewable. A brief that is silent on it will be read as if the trigger exists, and the gap will surface at the point where it is most expensive to fix.

Video briefs, answered

How do you write a video brief for an agency?: Commit five decisions to paper before you approach anyone: the funnel stage the asset serves, the single action you want a viewer to take, the martech trigger that fires behind that action, the contact record the view should touch, and the pipeline outcome the asset will be judged on. Everything else in a brief — tone, references, runtime, delivery formats — is production detail that a good agency can propose. Those five are the ones only you can answer, and an agency that is not given them will substitute its own.
What should a video brief include?: The five decisions above, plus the practical constraints: budget range, deadline, who approves, where the asset will live, and what already exists that it must sit alongside. A useful test is whether a stranger could read the brief and say what would count as success. If success is only describable as “a great film”, the brief is a mood board with a budget attached.
How long should a video brief be?: One to two pages. Length is not the quality signal — decisiveness is. A twelve-page brief that never names a funnel stage gives a production company less to work with than a single page that does. The 60-Second Brief is named for the test: if the five decisions cannot be stated aloud in sixty seconds, they have not been made yet.
Who should write the video brief, marketing or the agency?: Marketing writes the five decisions; the agency writes the treatment. That division is the point. The decisions are about your funnel, your stack and your pipeline, which the agency cannot see. The treatment is about story, craft and execution, which is what you are paying them for. Briefs go wrong most often when the client writes the treatment and leaves the decisions blank.
What is the most common mistake in a video brief?: Specifying the deliverable instead of the job. “We need a two-minute brand film” is a purchase order, not a brief — it names a format and a runtime and leaves the funnel stage, the viewer action and the success measure undefined. The asset then gets judged on whether people liked it, because nobody wrote down anything else it could be judged on.
Should the brief specify a budget?: Yes, as a range. Withholding it in the hope of a cheaper quote produces proposals scoped against a guess, which wastes both sides’ time and makes the responses incomparable. A range lets a production company tell you honestly what is achievable inside it, and lets you check the number against the acquisition cost the asset is meant to offset before anyone starts.

Related: the playbooks publish the 60-Second Brief as an artefact, how to choose a video production company covers vetting the vendor you send it to, the Buyer's Guide covers the commissioning decision behind it, and why brand films don't generate leads is what happens when the five decisions are skipped.