How do you measure video marketing ROI in B2B?

You measure video marketing ROI by dividing the acquisition cost a video offsets over its useful life by what it cost to produce. That requires three things to be true before the arithmetic means anything: the video was commissioned against a defined funnel stage, a view can be resolved to a known contact, and the brief tied the asset to a pipeline outcome. Without those, you can compute a number, but it is not measuring the video.

What is a good ROI for video marketing?

A corporate video that offsets less acquisition cost than it cost to produce is a brand expense, not an investment. Between one and three times its cost, it pays for itself and can be instrumented further. Above three times, it is a pipeline asset and the pattern is worth repeating. These are the tiers the CAC-to-Frame Ratio uses, and the ratio is deliberately conservative: it counts only acquisition cost you can trace.

Why are views and watch time not a measure of ROI?

Views and watch time measure consumption, not contribution. They have no denominator in money and no link to a contact record, so they cannot be divided into spend or joined to a deal. They are useful for diagnosing whether an asset holds attention. They cannot answer whether it paid for itself, and presenting them as though they can is what erodes finance’s trust in the whole line item.

How long should you wait before measuring a video’s return?

Measure the instrumentation immediately and the return over the asset’s useful life. Whether a view resolves to a known contact is observable within days of launch, and if it does not, no amount of waiting will fix it. The acquisition-cost offset only accumulates across a full buying cycle, so judging a film on its first quarter usually measures the sales cycle rather than the film.

Can you measure video ROI without a full attribution platform?

Yes, for one asset at a time. Instrument a single video end to end — a defined stage, a tracked next step, one automation that fires on view — and you can trace that asset without new software. What a platform buys you is doing this at portfolio scale. What it cannot buy you is a chain that was never wired in the first place.

What should you report to finance?

Report the acquisition cost offset, the number of traceable deals it rests on, and the share of video-influenced pipeline you cannot yet trace. The last figure is the one that builds credibility: naming the untraceable portion is what separates a measurement from a claim, and it is usually the largest number on the page.

Field guide

How to measure video marketing ROI

The number finance asks for, and the one you have.

By Paul Joseph · Updated 15 Aug 2026

You measure video marketing ROI by dividing the acquisition cost an asset offsets over its useful life by what it cost to produce — a number in money over a number in money. Before that division means anything, three links have to be intact: Pipeline (the goal the asset serves), Stack (the plumbing that carries a viewer to a known contact), and Craft (the brief that gives the asset a job). Views, completion rates and comment sentiment measure consumption, not return, and cannot be divided into spend. This page is how the number is built, what has to be true before it holds, and a full worked example you can run against your own figures.

Start here

If you want the diagnosis before the method, the Dark-Funnel Video Audit scores whether your spend is traceable at all across Pipeline, Stack, and Craft — two minutes, ten questions, no login.

Which numbers belong in the report in the first place, and which describe distribution rather than outcome, is set out in the video marketing metrics that predict pipeline.

Why the usual video metrics cannot answer the question

Views, watch time, completion rate, engagement rate, share of voice: every one of these measures consumption. They tell you whether an asset held attention, which is a real and useful thing to know. What none of them has is a denominator in money or a join to a contact record. You cannot divide a view count into a production budget and get a return, and you cannot join a view count to a closed deal without an identity in between.

This is why video is structurally harder to defend than paid search. A search click arrives with an identity, a cost, and a session you can follow. A video view arrives with none of those unless someone deliberately wired them in. The gap between the two is not a craft problem. It is a plumbing problem, and it has a name: the dark funnel, the untracked buyer activity between a viewed asset and a known contact.

Two established findings explain why that gap is so wide in B2B specifically. The 95-5 rule — Professor John Dawes of the Ehrenberg-Bass Institute, for the LinkedIn B2B Institute (2021) — holds that roughly 95% of business buyers are out-of-market at any given moment. Most of the people who watch your film cannot act on it now, and will act months later through a channel that carries no memory of the video. Gartner's B2B buying research (2024) puts the share of the buying journey spent with sales representatives at around 17%, which means the great majority of the decision forms where you are not present to observe it. Video is doing work in that space. Standard analytics is not built to see it.

What a return actually requires

A return is a ratio. For corporate video, the honest form of it is the acquisition cost the asset offsets over its useful life, divided by what it cost to produce. Frame to Funnel calls this the CAC-to-Frame Ratio, and the reason it is expressed against customer acquisition cost rather than revenue is practical: finance already funds acquisition cost, already has a number for it, and already reviews it line by line. Expressing a film as a share of the CAC it offsets puts it in a column that exists, instead of asking for a new one.

Three inputs are required, and each is a place the measurement usually fails.

01 · The production cost

The easiest input, and the one most often understated. It is not the invoice. It is the invoice plus the internal time to brief, review, and approve — usually the larger of the two on a corporate film, and always the part finance already knows about.
02 · The useful life

How long the asset stays in market before the product, the positioning, or the people in it date it. A film judged on one quarter is usually being judged on the sales cycle rather than on itself.
03 · The traceable offset

The acquisition cost avoided on deals where a video touch can actually be evidenced. This is the input that fails, and it fails upstream of the spreadsheet — in whether the view ever resolved to a contact.

The three links that must be intact first

Before any of those inputs can be trusted, the chain that carries a viewer to a tracked contact has to be unbroken. The Frame to Funnel Method names the three links, and a measurement exercise that skips them produces a number with nothing underneath it.

Pipeline — the goal. The funnel stage the video serves was defined before the camera rolled. Without it there is no outcome to measure the view against, and any number you produce is chosen after the fact to suit the result.
Stack — the plumbing. A view can be resolved to a known contact and a tracked next step. This is the link that decides whether measurement is possible at all.
Craft — the fuel. The brief tied the asset to a pipeline outcome, so the film is being asked to produce a result it was actually designed to reach.

The order matters. Fixing craft on a film whose views resolve to nobody improves the film and changes nothing about the measurement. Fixing the stack on a film with no defined stage gives you clean data about an undefined goal. The links have to be repaired in sequence, and the weakest one sets the ceiling on everything above it.

A worked example, with the assumptions stated

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.

Worked example inputs for a CAC-to-Frame calculation
Blended customer acquisition cost	$6,000
Brand film — agency invoice	$34,000
Internal brief, review and approval time	$11,000
Total production cost	$45,000
Useful life	24 months
Deals closed in that window with an evidenced video touch	60
Acquisition cost avoided per such deal	$900 (15% of blended CAC)

The offset is 60 × $900 = $54,000. Divided by the $45,000 production cost, the CAC-to-Frame Ratio is 1.2×. The film paid for itself with a little to spare — in the calculator's terms, justified spend rather than a pipeline asset. Nobody is getting promoted for 1.2×, but nobody is defending a write-off either.

Now change one input. Suppose the stack can only evidence a video touch on 20 of those deals rather than 60 — not because the film influenced fewer, but because the other 40 arrived through direct visits and referrals that carried no record of it. The offset falls to $18,000, and the ratio to 0.4×. The film is now, on paper, a loss.

Nothing about the film changed. Nothing about its actual influence changed. The only thing that changed was how much of that influence the plumbing could see. This is the single most important thing to understand about measuring video: most of the variance in reported video ROI is variance in instrumentation, not in the video. A team with a wired stack and a mediocre film will report a better return than a team with a broken stack and an excellent one, and both reports will be honest.

Which is why the measurement question has to be asked before the commissioning question, not after the invoice. Run your own numbers in the CAC-to-Frame calculator, then treat the gap between the deals you believe video influenced and the deals you can evidence as your actual project.

What to measure at each stage

A single return figure for "video" as a category is not useful, because a film built for awareness and a testimonial built for the decision stage are being asked to do different jobs. The Pipeline Video Framework maps each asset to the buyer question it answers, and each stage has a measure that suits it.

Awareness. Measure whether reach converts to identity at all — the share of views that produce a resolvable contact, however weakly. Return here is measured in known audience, not in deals, and the 95-5 rule is the reason: you are paying to be remembered by people who cannot buy yet.
Consideration. Measure progression: the share of viewers who take the tracked next step the asset was built to prompt. This is the stage where instrumentation pays for itself fastest.
Decision. Measure influence on close rate and cycle length for deals where the asset was viewed by someone on the buying committee. This is the stage where CAC offset is most defensible, and the stage most video libraries have nothing in.
Advocacy. Measure reuse — whether customers and sellers actually deploy the asset. An advocacy film nobody sends is a cost regardless of how it tests.

What to start on Monday

Do not begin by measuring the library. Begin by instrumenting one asset end to end: pick the video closest to a buying decision, define the stage it serves, give the viewer one tracked next step, and fire one automation when that step is taken. Then watch what the stack can and cannot see for a full buying cycle. One instrumented asset teaches you more about your measurement gap than a quarter of reporting on ten untracked ones, and it produces the first number in this entire exercise that will survive a question from finance.

The episode The Pipeline Video Framework works through how a library gets maldistributed across stages in the first place, and Why your six-figure brand film generated zero leads is the teardown of what this looks like when it fails completely.

The four mistakes that make the number worthless

Each of these produces a figure that looks like a return and is not one. They are worth naming because all four survive review — nobody in the meeting can see what is wrong with them from the slide.

Crediting the whole deal to the asset. A film that was one of eleven touches did not produce the revenue. Attributing full deal value to it produces a spectacular ratio and destroys the credibility of every future number you present. Measuring against acquisition cost offset rather than revenue avoids this by construction, because you are claiming a saving, not a sale.
Counting only the invoice. Omitting internal brief, review, and approval time can understate the denominator substantially on a stakeholder-heavy corporate film, which inflates the ratio by the same proportion. If the number is going to finance, they will find this.
Measuring the quarter instead of the life. Judging a 24-month asset on 90 days measures your sales cycle. It is the most common reason a perfectly sound film is written off eight months before it would have paid back.
Silently dropping the untraceable. Reporting only evidenced influence, with no statement of what could not be traced, presents a floor as though it were a total. It understates video and, worse, hides the instrumentation problem that is actually costing you the difference.

What a quarterly video review should contain

Most video reporting is a highlights reel with numbers attached. A review that survives scrutiny and actually informs the next commission is a short, dull document with five things in it.

1. The asset register. Every live video, the funnel stage it serves, and the date it goes out of date. This alone surfaces the stage with nothing in it, which is usually the finding that changes next quarter's budget.
2. Instrumentation status per asset. Tracked next step, identity resolution, automation — present or absent. Three columns, ticks and crosses. The crosses are the work.
3. Evidenced influence. Deals with a recorded video touch, and the acquisition cost offset that implies, stated with the assumption you used for the offset rate.
4. The traceability gap. Deals where sales or the buyer reported video influence that the stack did not record. This is the number to watch quarter over quarter, and it should be falling.
5. One decision. What is being commissioned, retired, or rewired as a result. A review that produces no decision is a report, and reports do not change ratios.

The discipline that makes this work is stating your assumptions in the document rather than defending them in the meeting. The offset rate is an assumption. The useful life is an assumption. Written down, they become things a CFO can argue with productively — and an assumption someone argues down is still a shared model, which is more than most video reporting ever achieves.

When the number will not come

Sometimes the honest answer to "what did the video return" is that it cannot yet be established, and the useful response is to say so with a date attached rather than to manufacture a figure. Three situations produce this, and each has a different remedy.

The asset is too young. If the film has been live for less than one full buying cycle, there is no return to measure yet — only instrumentation to confirm. Report the instrumentation status and name the date the return question becomes answerable.

The chain was never wired. If views cannot resolve to contacts, no analysis will recover the answer retrospectively, and it is better to say that plainly than to model around it. The remedy is to instrument the next asset and treat this one as the cost of learning that.

The volume is too low. An asset touching a handful of deals produces a ratio that swings wildly on one closure. Report the direction and the sample size, and resist the temptation to annualise from three data points — that is where video reporting most often loses its audience for good.

Measuring video ROI, answered

How do you measure video marketing ROI in B2B?: You measure video marketing ROI by dividing the acquisition cost a video offsets over its useful life by what it cost to produce. That requires three things to be true before the arithmetic means anything: the video was commissioned against a defined funnel stage, a view can be resolved to a known contact, and the brief tied the asset to a pipeline outcome. Without those, you can compute a number, but it is not measuring the video.
What is a good ROI for video marketing?: A corporate video that offsets less acquisition cost than it cost to produce is a brand expense, not an investment. Between one and three times its cost, it pays for itself and can be instrumented further. Above three times, it is a pipeline asset and the pattern is worth repeating. These are the tiers the CAC-to-Frame Ratio uses, and the ratio is deliberately conservative: it counts only acquisition cost you can trace.
Why are views and watch time not a measure of ROI?: Views and watch time measure consumption, not contribution. They have no denominator in money and no link to a contact record, so they cannot be divided into spend or joined to a deal. They are useful for diagnosing whether an asset holds attention. They cannot answer whether it paid for itself, and presenting them as though they can is what erodes finance’s trust in the whole line item.
How long should you wait before measuring a video’s return?: Measure the instrumentation immediately and the return over the asset’s useful life. Whether a view resolves to a known contact is observable within days of launch, and if it does not, no amount of waiting will fix it. The acquisition-cost offset only accumulates across a full buying cycle, so judging a film on its first quarter usually measures the sales cycle rather than the film.
Can you measure video ROI without a full attribution platform?: Yes, for one asset at a time. Instrument a single video end to end — a defined stage, a tracked next step, one automation that fires on view — and you can trace that asset without new software. What a platform buys you is doing this at portfolio scale. What it cannot buy you is a chain that was never wired in the first place.
What should you report to finance?: Report the acquisition cost offset, the number of traceable deals it rests on, and the share of video-influenced pipeline you cannot yet trace. The last figure is the one that builds credibility: naming the untraceable portion is what separates a measurement from a claim, and it is usually the largest number on the page.

Related: video attribution covers how a view is resolved to a contact in the first place, without which none of this arithmetic runs. The dark funnel explains why the traceable offset is always smaller than the real one, what a corporate video should cost sets the denominator in the ratio, and B2B video marketing is the wider system all of it sits inside. Does video lower customer acquisition cost? takes the prior question — whether the offset exists at all — and the 95-5 rule and your video budget covers why the measurable half of the return is never the whole of it. The glossary defines every term used here, and the Corporate Video Buyer's Guide covers how to commission the next asset so it is measurable before it is made.