How do you defend a video marketing budget?

By changing the unit the line is reported in, conceding the untraceable share before anyone asks for it, and tying the spend to a number somebody outside marketing is already accountable for. A defence built on views, engagement or output volume loses even when the programme is working, because those are marketing’s units and the person deciding is working in payback, contribution and cost avoided. The strongest position in the room is not the best result — it is the most precisely stated one, including about what you cannot yet show.

Why is video the first marketing line to get cut?

Because of three properties that have nothing to do with whether it works. It is lumpy, so it appears as a large discrete number rather than a rate. It is discretionary, so nothing breaks next week if it stops. And it is visible, so everyone in the room has an opinion about the output in a way they do not about a data platform renewal. A line with those three properties is the path of least resistance for anyone who has been told to find a percentage.

What should a CMO bring to a video budget review?

Three things and no deck: the asset register with each asset’s stage and instrumentation status, the evidenced influence with the offset assumption written into the document rather than defended aloud, and the traceability gap stated as a quantity that is falling. Anything else you bring will be read as padding, and padding invites the question you cannot answer. Bring the decision you want out of the meeting too, phrased as one sentence.

What do you say when you cannot prove video ROI yet?

Say it plainly, then stop defending and start trading. "We cannot currently trace a view to a contact, so any number I give you is an estimate. Here is what it costs to fix that, and here is the date you will get a real number." A finance partner will fund an instrumented pilot far more readily than an unproven programme at full scale, because the pilot has a defined end and a defined output. Arguing from conviction against someone who deals in evidence loses on their ground every time.

Which video metrics should a CMO never take into a board meeting?

Impressions, view counts, average watch time and asset volume. The first two measure distribution and rise when you spend more on it, which makes them circular. Average watch time collapses a split distribution into a figure that describes nobody. Asset volume measures activity, and reporting activity to someone who allocates capital is how a line gets read as a cost centre. Keep all four for diagnosis; they belong in the team’s working file, not in front of the person who can cut you.

Who actually decides whether the video budget survives?

Rarely the person who challenges it. In most reviews the CFO or the board asks a question, marketing cannot answer it in the room, and the line is quietly reduced in a later document that nobody argues with. The decision is made by the silence, not by the challenge. That is why the useful preparation is not a stronger argument but a shorter list of questions you cannot answer — and a stated plan for each one that remains.

Field guide · for marketing leaders

How to defend a video marketing budget

The line rarely dies of a bad number. It dies of a silence.

By Paul Joseph · Updated 26 Aug 2026

Three moves carry a video budget through a review. Report the line in the units of the person who can cut it, which are payback, contribution and cost avoided rather than views and engagement. Concede the share you cannot trace before anyone asks for it, stated as a quantity. And attach the spend to a number somebody outside marketing is already accountable for, so cutting it becomes their problem as well as yours. None of the three is about the video.

This page is about the meeting. Measuring video marketing ROI is about the document you take into it — the formula, the assumptions, and what a quarterly review should contain. Get that one right first; this one is what happens when it meets somebody who has been told to find fifteen per cent.

What is actually being challenged

Almost never the video. What is under pressure is marketing's claim on capital as a whole, and video is simply the easiest place to apply it. Three properties make it so, and none of them is performance.

It is lumpy. A commission arrives as one large discrete number rather than a monthly rate, so it is legible as a thing to remove. A platform costing the same across the year is invisible by comparison, because nobody has to approve it twice.
It is discretionary. Stop it and nothing breaks on Monday. Stop the CRM renewal and sales cannot work. Whether that difference reflects relative value is beside the point — it reflects relative consequence, and consequence is what a person under pressure optimises for.
It is visible. Everyone in the room has watched the film and has a view about it. Nobody has an aesthetic opinion about a data warehouse. Visibility feels like an advantage when the work is admired and becomes a liability the moment the conversation turns to cost.

The practical consequence is that arguing about the quality of the work is arguing about the wrong thing. A film the board liked is not evidence in a capital-allocation conversation, and offering it as evidence signals that you do not have any of the other kind.

Report in the other person's units

A view count is not a weak number. It is a number in the wrong currency, and currency conversion is the whole job. The person deciding works in payback period, contribution, cost avoided and risk. Anything you hand them that is not already in one of those has to be converted by them, in their head, during your presentation — and the conversion rate they apply will be less generous than yours.

Marketing's reporting units converted into the units used in a budget review
What marketing reports	What it converts to	Why the second survives the room
Views, impressions	Nothing. Leave them out	They rise when you spend more on distribution, which makes the argument circular
Engagement rate	Next-step rate on the one action the asset asked for	It names a behaviour with a business consequence rather than an interaction
Assets produced	Cost per influenced deal	Output volume reads as activity; activity is what cost centres report
Brand lift, sentiment	Acquisition cost avoided, with the lag stated	It concedes the timescale instead of being caught by it

Which video measures carry information works through the right-hand column properly. The point here is narrower: the conversion has to happen before the meeting, in your document, or it happens during the meeting, in their head.

Concede first, and concede precisely

Every video programme has influence it cannot evidence. The instinct is to keep quiet about it and hope the question does not come. The question always comes, and answering it under pressure is the worst position to answer it from — you will either overclaim, which costs you the rest of the meeting, or concede vaguely, which sounds like the whole number is soft.

Volunteer it instead, early, as a quantity. "Eleven deals last quarter had a recorded video touch. Sales reported influence on nineteen. The gap of eight is what we cannot currently evidence, and it was fourteen two quarters ago." The content of that sentence matters less than when it is said. Disclosed by you in the first two minutes, it is a demonstration that you know where your own weakness is. Extracted from you in the twentieth, it is a discovery — and a reviewer who has just discovered one soft number will reasonably assume there are others.

The mechanics of holding a line while that gap is still open — reporting both numbers labelled, committing to a reduction with a date, and what buyer verbatims can and cannot carry — are worked through in the dark funnel, which is also where the gap comes from. The useful thing to be able to say in the room is that it is a plumbing problem rather than a marketing one, because plumbing problems have quotes attached and a date by which they close.

The three defences that lose the room

The benchmark. "Our engagement is above the industry average." Nobody in the room is accountable for the industry average, and citing one invites the obvious reply that the industry is also not growing. A benchmark is a comparison, and you are being asked for a consequence.
The single anecdote. "A customer told us on a call that the film is why they shortlisted us." True, and genuinely evidence, and useless on its own — one data point, reported by an interested party, offered against a number. The distinction worth holding is between one story told from memory and fifteen buyer verbatims collected the same way every quarter: the second is a small qualitative sample and belongs in the document, the first is a reflex and belongs nowhere near the meeting.
The volume. "We produced forty assets this year." This is the most dangerous of the three, because it sounds like productivity and reads as cost. It also invites exactly the wrong follow-up question, which is whether forty could be done for less — and now you are negotiating unit price on work whose value you have not established.

A worked example: one programme, two framings

Illustrative, not a benchmark. The same twelve months of activity, the same underlying facts, presented two ways. Nothing has been added to the second column — the difference is entirely which facts were selected and what they were converted into.

The same video programme reported in marketing's units and in finance's units
As usually presented	As it survives
$140,000 spent across nine assets	$140,000 against 31 deals carrying a recorded video touch — $4,500 per influenced deal
410,000 views, engagement up 22% year on year	1,900 next steps taken from the one action each asset asked for; the two consideration assets account for 71% of them
Strong feedback from sales and the board	Sales reported influence on 47 deals; 31 are evidenced. The 16-deal gap is the instrumentation backlog, down from 29
Recommend maintaining the budget	Recommend holding the budget and moving $30,000 of it from awareness to consideration, on the 71% above

One caution about that first figure, because it is the one a reader converts wrongly. $4,500 per influenced deal is not a saving against the blended acquisition cost, and presenting it as one will be caught: those 31 deals carried their full acquisition cost as well. It is a unit price for influence, useful because it is comparable against the same figure next quarter and against other channels running the same calculation. Offer it as a ratio to move, never as a cost avoided, and the person opposite has nothing to correct.

Two things are doing the work in the right-hand column, and neither is a better result. The first is that every figure is attached to a decision rather than to a period. The second is that the weakest number in the set — the sixteen deals that cannot be evidenced — is stated by you, with a direction of travel, rather than discovered by someone else. The recommendation is also specific enough to be argued with, which sounds like a risk and is the opposite: a proposal that can be amended survives a meeting that a proposal to "maintain" does not.

The $4,500-per-influenced-deal figure rests on an assumption about how much of each deal the asset actually moved, and that assumption belongs in the document in plain sight. The CAC-to-Frame calculator works out what a given asset has to earn back; the discipline it enforces is stating the offset rate rather than burying it, so the reviewer argues with the assumption instead of with the conclusion.

When you genuinely cannot prove it yet

Sometimes the honest position is that a view cannot currently reach a known contact, in which case every number you could offer is an estimate and a competent finance partner will identify that within two questions. Do not defend from there. Trade instead.

The trade is specific: name what it costs to instrument one asset end to end, name the date a real number arrives, and ask for the programme to be held at its current level until then rather than argued about now. This works because it converts an open-ended commitment into a bounded one with a deliverable, which is a shape a finance function is built to approve. It also concedes the thing they were going to establish anyway, which spends nothing and buys credibility.

Instrumenting one asset is genuinely cheap and takes days rather than quarters — the Dark-Funnel Video Audit scores where your own chain breaks in two minutes and names the link to wire first, which is enough to put a number and a date on the trade. If the assets exist, produce nothing, and the instrumentation is already fine, the problem is elsewhere: when a B2B video strategy is not working separates the four causes.

Who actually decides

Rarely the person who asks the hard question. In most organisations the challenge is raised in the room, the answer is incomplete, nobody argues, and the line is reduced weeks later in a document that circulates without a meeting. The decision is made by the silence rather than by the challenge, which is why preparing a stronger argument is the wrong preparation. The right preparation is a shorter list of questions you cannot answer.

Write that list down before the review — honestly, including the ones you hope will not come up. For each, decide whether the answer is a number you can go and get, a concession you should volunteer first, or a trade you should propose. A question with one of those three attached to it is no longer a threat. A question with nothing attached is the one that ends the line.

One more thing worth doing early rather than in the room: find out whose number your programme is nearest to — pipeline coverage, cycle time, cost per acquisition — and get that person to agree, in advance, that video touches it. A line that only marketing defends is a marketing line. A line that the person accountable for pipeline would rather not lose is a different object entirely.

Defending a video budget, answered

How do you defend a video marketing budget?: By changing the unit the line is reported in, conceding the untraceable share before anyone asks for it, and tying the spend to a number somebody outside marketing is already accountable for. A defence built on views, engagement or output volume loses even when the programme is working, because those are marketing’s units and the person deciding is working in payback, contribution and cost avoided. The strongest position in the room is not the best result — it is the most precisely stated one, including about what you cannot yet show.
Why is video the first marketing line to get cut?: Because of three properties that have nothing to do with whether it works. It is lumpy, so it appears as a large discrete number rather than a rate. It is discretionary, so nothing breaks next week if it stops. And it is visible, so everyone in the room has an opinion about the output in a way they do not about a data platform renewal. A line with those three properties is the path of least resistance for anyone who has been told to find a percentage.
What should a CMO bring to a video budget review?: Three things and no deck: the asset register with each asset’s stage and instrumentation status, the evidenced influence with the offset assumption written into the document rather than defended aloud, and the traceability gap stated as a quantity that is falling. Anything else you bring will be read as padding, and padding invites the question you cannot answer. Bring the decision you want out of the meeting too, phrased as one sentence.
What do you say when you cannot prove video ROI yet?: Say it plainly, then stop defending and start trading. "We cannot currently trace a view to a contact, so any number I give you is an estimate. Here is what it costs to fix that, and here is the date you will get a real number." A finance partner will fund an instrumented pilot far more readily than an unproven programme at full scale, because the pilot has a defined end and a defined output. Arguing from conviction against someone who deals in evidence loses on their ground every time.
Which video metrics should a CMO never take into a board meeting?: Impressions, view counts, average watch time and asset volume. The first two measure distribution and rise when you spend more on it, which makes them circular. Average watch time collapses a split distribution into a figure that describes nobody. Asset volume measures activity, and reporting activity to someone who allocates capital is how a line gets read as a cost centre. Keep all four for diagnosis; they belong in the team’s working file, not in front of the person who can cut you.
Who actually decides whether the video budget survives?: Rarely the person who challenges it. In most reviews the CFO or the board asks a question, marketing cannot answer it in the room, and the line is quietly reduced in a later document that nobody argues with. The decision is made by the silence, not by the challenge. That is why the useful preparation is not a stronger argument but a shorter list of questions you cannot answer — and a stated plan for each one that remains.

Related: setting a video marketing budget is the number this review is about, B2B video by funnel stage is where a held budget should be moved to, and the glossary defines every term used here.