What percentage of a marketing budget should go to video?

This is the question most buyers arrive with and it has no useful answer. Two companies with identical revenue can justify budgets an order of magnitude apart, because the variable is not company size — it is how much of the buying decision happens in channels video reaches, and whether the stack can resolve a view to a contact. A company that cannot trace a view should spend less until it can, whatever the percentage says.

Should the video budget sit with brand or demand generation?

Wherever accountability for the outcome sits. The common failure is splitting them: brand commissions the asset, demand generation is measured on the pipeline it was supposed to produce, and neither owns the gap between. One budget line with one owner produces better assets than two lines with shared credit, almost regardless of which team holds it.

How much of the video budget is not production cost?

More than most budgets show. Briefing, review cycles, stakeholder approvals and post-launch distribution consume internal hours that rarely appear against the video line, and they are the costs most inflated by an unclear brief — ambiguity gets resolved through review rounds. A budget that counts only the production invoice is understating the real figure, which matters because the real figure is the denominator in every return calculation.

How do you defend a video budget to finance?

Express it in the units finance already funds. Acquisition cost is a line that exists, has a number attached and is reviewed regularly. A film presented as a share of the acquisition cost it offsets sits inside that conversation. A film presented as a production budget asks for a new line, in a language the reviewer does not use, and is cut first when the quarter tightens.

When should a company spend less on video, not more?

When a view cannot be resolved to a known contact. Until that link works, additional spend increases the volume of untraceable influence rather than the volume of pipeline, and the programme cannot be evaluated at any budget level. Fixing the tracking layer is cheaper than another film and is the precondition for the next one being measurable.

Field guide

How to set a B2B video marketing budget

There is no correct percentage.

By Paul Joseph · Updated 23 Aug 2026

Set a B2B video marketing budget from the acquisition cost the assets have to offset, not from a share of marketing spend. Estimate the deals video can plausibly influence over the assets’ useful life, multiply by the acquisition cost each influenced deal avoids, and treat that figure as the ceiling. Any percentage you are quoted — of revenue, of marketing budget, of anything — is describing a funnel that is not yours. This page covers why that is, how to run the calculation, the costs that never reach the video line, and how to defend the number to someone whose job is to cut it.

If you read one thing

A budget is not a size problem, it is a traceability problem. Until a view can be resolved to a known contact, additional spend buys more untracked influence — which is real, and which you cannot evidence. That gap has a name, the dark funnel, and a two-minute diagnostic, the Dark-Funnel Video Audit.

Why the percentage question has no answer

The question arrives in a recognisable form. Someone in the leadership meeting asks what other companies of our size spend on video, an analyst report gets circulated, and a number is adopted because it came from outside the building. It survives review precisely because nobody in the room owns it.

The problem is not that such figures are wrong. It is that they are answers to a different question. A percentage benchmark describes the average of a population you are not in: different deal sizes, different sales cycles, different buying committees, different degrees of instrumentation. Two companies with identical revenue can justify video budgets an order of magnitude apart without either being mismanaged, because the variable that matters is not their size.

The variable that matters is how much of the decision happens where video can reach. Gartner’s B2B buying research (2024) puts the share of the buying journey spent with sales representatives at around 17%. The rest is self-directed — peers, search, content consumed without anyone identifying themselves. A company whose buyers do most of their evaluating in that unobserved space has a stronger case for video spend than one whose deals are won in a room. Neither case is expressed by a percentage of revenue.

The second variable is whether you can see any of it. A budget spent by an organisation that cannot resolve a view to a contact buys influence it will never be able to evidence. That is not an argument for spending nothing. It is an argument for fixing the tracking layer first, because the same money spent after that fix produces a programme that can be evaluated, and the same money spent before it produces one that cannot.

What the budget is actually buying

Before sizing anything, be specific about what the money is for. A buying committee moves through four questions, and the Pipeline Video Framework maps an asset type to each. Most libraries are heavily weighted toward the first and empty at the third, which is the one nearest revenue.

Is this a problem we have? The cheapest assets to brief and the safest to approve, which is why budgets drift here. They are also the hardest to attribute, because the viewer is furthest from a decision.
How would this work for someone like us? Demonstrations and walkthroughs specific enough that a viewer recognises their own situation. Expensive to make well, and the first thing cut when the budget is set by percentage rather than by stage.
Can this vendor be defended internally? The stage closest to revenue and the one most libraries have nothing in. A champion forwarding an asset to a sceptical finance lead is doing the work your sales team cannot be in the room for.
Was this the right decision? Post-sale assets that reduce churn and produce the references the next deal depends on. Almost never funded from the video line, and almost always attributed to customer success instead.

Read your own library against those four before setting a number. A budget that funds a fifth awareness film while the third question goes unanswered is not too small. It is pointed at the wrong stage, and increasing it will produce more of the same.

Sizing the budget from acquisition cost

The defensible method inverts the usual order. Instead of setting a budget and asking what it can buy, establish what the assets would have to be worth to justify a given spend, then decide whether that is plausible. This is the CAC-to-Frame Ratio applied to a programme rather than a single film.

The figures below are illustrative arithmetic, not benchmarks. Substitute your own — the point is the shape of the calculation and where it breaks, not these numbers.

Illustrative inputs for sizing an annual video programme budget
Blended customer acquisition cost	$6,000
Deals closed per year	400
Share with an evidenced video touch	25% (100 deals)
Acquisition cost avoided per such deal	$900 (15% of blended CAC)
Annual acquisition cost offset	$90,000
Useful life of the assets	24 months
Total offset over that life	$180,000

An annual programme costing $90,000 against a two-year offset of $180,000 runs at roughly 2×. That is a defensible position: the assets pay for themselves inside their life and leave margin for the estimate being optimistic. At $180,000 the programme breaks even only if every assumption holds, which none of them will. At $45,000 there is room to spend more, and the constraint is production capacity rather than economics.

Two of those inputs are where the argument actually happens. The share with an evidenced video touch is not a market statistic, it is a measurement of your own instrumentation — and if you cannot produce it, that is the finding, not a gap to fill with an assumption. Video attribution is how that number gets built. The acquisition cost avoided per deal is a judgement about contribution, and it should be set conservatively enough that you would defend it to a sceptic, because you will have to.

Run the calculation twice: once with the numbers you can evidence, and once with the numbers you believe. The gap between them is the size of your attribution problem, expressed in money. It is usually a more persuasive argument for fixing the stack than anything said about the stack itself.

The costs that never reach the video line

A budget built from production invoices understates the real figure, sometimes substantially. That matters for one specific reason: the understated figure is the denominator in every return calculation you will later run, so the programme looks more efficient than it is and the next budget is set on a false baseline.

Internal time. Briefing, review cycles, stakeholder approvals. Routinely rivals or exceeds the production invoice, almost never appears against the video line, and is the cost most inflated by an unclear brief — ambiguity is resolved through review rounds, and review rounds are salaried hours.
Distribution. The paid promotion, landing pages and email sequences that decide whether anyone sees the asset. Often held in a separate budget, which is how a film gets made and then distributed on whatever is left over.
The stack work. Wiring a view to a tracked contact is engineering effort, and it is the thing that makes the rest measurable. Budget it explicitly or it will not happen.
Maintenance. A product changes, a claim expires, a customer in a testimonial leaves. Assets decay, and an unbudgeted asset decays until someone notices it is embarrassing.
The opportunity cost of the approval chain. Not a line item, but real: a film that takes fourteen weeks to approve has consumed a quarter of its own useful life before it launched.

What a corporate video should cost covers the five drivers that move a single quote. This page is about the programme, but the same principle governs both: a number with no specification behind it is a guess, whether it is one film or a year of them.

Splitting the budget across the funnel

Once the total is set, the allocation question is where most programmes quietly fail. The default split is not chosen; it emerges from what is easiest to get approved, which is why awareness work accumulates. Two rules keep the split honest.

Fund the stage your library cannot currently answer. Not the stage that produces the best showreel. If a champion has nothing to forward to a sceptical CFO, that is the gap, and it is nearer revenue than another brand film.

Reserve for the plumbing before the fuel. If the tracking layer is not built, take the cost of building it off the top rather than hoping it fits later. A smaller programme that can be measured outperforms a larger one that cannot, because the smaller one tells you what to do next year.

Those two rules govern allocation by funnel stage. The other split — how much goes to buyers who are in market today versus the far larger number who are not — is a separate argument with its own arithmetic, and the 95-5 rule and your video budget works it through properly. The short version for sizing purposes: only the conversion half will produce evidence inside a planning cycle, so a total set on this year's provable return will systematically underfund the half that compounds.

Defending the number

A video budget is cut for a predictable reason: it is presented in units nobody else in the review uses. Every other line is expressed as a cost of acquiring or retaining customers. Video arrives as a production budget, which asks the reviewer to accept a new category on trust, in a quarter when they are removing categories.

Present it the way the rest of the model is presented. The programme offsets a stated amount of acquisition cost over a stated life, at a stated ratio, with the assumptions written down and the weakest one named before anyone else finds it. That last part matters more than it sounds: naming your own weakest assumption is what separates a budget request from a pitch, and reviewers who cut pitches tend to fund requests.

Bring the version you can evidence and the version you believe, and be explicit that the difference is a measurement gap rather than a rounding one. A finance lead who is shown the honest smaller number and told exactly what it would take to make the larger one provable is being offered a decision. One who is shown only the larger number is being asked for faith.

When to spend less

Three conditions argue for reducing the budget rather than defending it, and recognising them early is worth more than any allocation model.

A view cannot be resolved to a contact. Additional spend buys additional untraceable influence. Fix the handoff, then spend. Why brand videos don’t generate leads is what this looks like from the outside, usually about six months later.

Nobody owns the asset after launch. An unowned asset stops being distributed within weeks. The cheapest improvement available to most programmes is not another film; it is naming someone accountable for the contribution of the ones already made.

The brief cannot name a funnel stage. If the commissioning conversation cannot say which buyer question the asset answers and what the viewer is meant to do next, the money is committed to a mood. The Corporate Video Buyer’s Guide covers that conversation.

There is a fourth case, and it is the most common one this page will be read in: the number is already fixed and you did not set it. Sizing arithmetic is still worth running, but its output changes. It stops being a budget request and becomes an allocation argument — the same calculation, applied to each candidate asset, ranks them by the acquisition cost each would offset for what it costs. Spend the fixed number on the two or three that clear the ratio rather than spreading it across six that do not, and record the ones you declined and why. Next year that record is the argument, and it is considerably harder to dismiss than a benchmark, because it is about your own funnel and the reviewer has seen the results.

Video budgets, answered

How much should a B2B company spend on video marketing?: There is no defensible percentage, and any figure quoted as an industry standard is describing someone else’s funnel. Size the budget from the acquisition cost the assets have to offset: estimate the deals video can plausibly influence over the assets’ useful life, multiply by the share of acquisition cost each influenced deal avoids, and treat that total as the ceiling on what the programme is worth spending. A budget derived that way survives a finance review; a percentage does not.
What percentage of a marketing budget should go to video?: This is the question most buyers arrive with and it has no useful answer. Two companies with identical revenue can justify budgets an order of magnitude apart, because the variable is not company size — it is how much of the buying decision happens in channels video reaches, and whether the stack can resolve a view to a contact. A company that cannot trace a view should spend less until it can, whatever the percentage says.
Should the video budget sit with brand or demand generation?: Wherever accountability for the outcome sits. The common failure is splitting them: brand commissions the asset, demand generation is measured on the pipeline it was supposed to produce, and neither owns the gap between. One budget line with one owner produces better assets than two lines with shared credit, almost regardless of which team holds it.
How much of the video budget is not production cost?: More than most budgets show. Briefing, review cycles, stakeholder approvals and post-launch distribution consume internal hours that rarely appear against the video line, and they are the costs most inflated by an unclear brief — ambiguity gets resolved through review rounds. A budget that counts only the production invoice is understating the real figure, which matters because the real figure is the denominator in every return calculation.
How do you defend a video budget to finance?: Express it in the units finance already funds. Acquisition cost is a line that exists, has a number attached and is reviewed regularly. A film presented as a share of the acquisition cost it offsets sits inside that conversation. A film presented as a production budget asks for a new line, in a language the reviewer does not use, and is cut first when the quarter tightens.
When should a company spend less on video, not more?: When a view cannot be resolved to a known contact. Until that link works, additional spend increases the volume of untraceable influence rather than the volume of pipeline, and the programme cannot be evaluated at any budget level. Fixing the tracking layer is cheaper than another film and is the precondition for the next one being measurable.

Related: B2B video marketing is the system this budget funds, measuring video marketing ROI covers the arithmetic once the year is spent, and what a corporate video should cost covers a single quote rather than a programme. The Does video lower customer acquisition cost? argues the premise the whole sizing method rests on, and the glossary defines every term used here.