How to Measure Video Marketing ROI Without Fooling Yourself

The video goes live on Monday.

By the end of the month, enquiries are up 30%.

Did the video work?

Maybe.

Or perhaps you also launched a new website, increased the advertising budget, attended a trade show and started a September promotion during the same four weeks.

The interesting part of video ROI is not producing a bigger percentage.

It is working out what changed, what the video plausibly contributed to, and whether that change was worth the money and effort involved.

Views, watch time and clicks can help explain what happened along the way.

They are not the return.

Start with the job you gave the video

Before opening Analytics, go back to the brief.

What was this film actually supposed to help with?

A customer case study might be intended to help prospective buyers feel more confident before speaking to sales.

A product demonstration might reduce the number of unsuitable enquiries by showing clearly what the product can and cannot do.

A recruitment film might help candidates decide whether the role suits them before applying.

A training film might reduce the amount of staff time spent repeating the same explanation.

Those are four different jobs.

They should not all be judged by views.

A useful starting sentence is:

If this video works, we should see…

For a sales film:

If this video works, we should see more qualified enquiries from the campaign and hear sales prospects referring to the case study.

For recruitment:

If this video works, candidates should understand the role better before interview, and fewer unsuitable applicants should reach the later stages.

For training:

If this video works, managers should spend less time repeatedly explaining this process and staff should still be able to complete it correctly.

Now there is something worth measuring.

Imagine a £4,000 customer case-study campaign

This is an illustrative example, not a Camera Guys package or a claim about a real client.

A business spends £4,000 on a customer case-study campaign.

That total includes:

• filming and editing;
• several shorter versions;
• page work;
• some paid distribution;
• and the internal time the business has chosen to include in the calculation.

During the next three months, the campaign is associated with eight qualified enquiries.

Three become customers.

Those three jobs bring in £24,000 of revenue.

At first glance, the video looks spectacular:

£24,000 revenue from £4,000 spend.

But revenue is not profit.

Suppose delivering those three jobs costs the business £14,000 in labour, materials and other relevant fulfilment costs.

That leaves £10,000 of contribution before the £4,000 campaign cost.

A simple ROI calculation would then be:

ROI = (£10,000 − £4,000) ÷ £4,000 × 100

ROI = 150%

That means the calculated return above the campaign cost is one and a half times the money invested.

It is a perfectly valid calculation if the cost basis and attributed value are valid.

And that “if” is where most of the real work sits.

Do not credit the video with every sale that happened afterwards

Three customers bought during the campaign.

Would all three have bought anyway?

Did one first meet the sales team at a trade show?

Did another arrive through Google and only see the video after the salesperson sent it?

Was the third directly driven by a paid post using a short extract from the case study?

Those are different journeys.

It would be dishonest to say:

“The video generated £24,000.”

when the evidence only tells you that the video was present somewhere in the process.

Marketing attribution is the process of assigning credit to different interactions. Google Analytics has attribution models for exactly this reason: different models can assign credit differently across the steps that lead to a key event. (support.google.com)

That reporting is useful.

It still does not turn a complicated buying decision into a controlled scientific experiment.

Treat attribution as evidence about contribution, not proof that one touchpoint caused the result.

Ask the sales team a question Analytics cannot answer

For a smaller B2B business, some of the most useful evidence may come from the people actually speaking to prospects.

Ask:

• Did the prospect mention the video?
• Did sales send the film before the meeting?
• Did the customer arrive understanding the service better?
• Did the film answer a question that normally takes ten minutes to explain?
• Did the customer story help overcome a particular concern?

Record those observations somewhere consistent.

Do not replace analytics with anecdotes.

Use both.

A prospect may arrive through search, watch the film, leave, speak to a colleague, come back directly two weeks later and then submit the form.

The website report may credit the final measurable interaction while the sales conversation reveals that the film played an important supporting role.

Neither view is the entire truth on its own.

Give campaign links names you can recognise later

If the film is being distributed through email, social posts, paid activity or partner links, give those routes consistent campaign tags where appropriate.

Google Analytics supports UTM parameters such as:

• utm_source
• utm_medium
• utm_campaign
• utm_content

so campaign traffic can be distinguished in acquisition reporting. Google recommends consistent naming because different spellings can split what should be one campaign into separate rows. (support.google.com)

For example:

utm_source=linkedin
utm_medium=organic_social
utm_campaign=customer_case_study
utm_content=30s_customer_clip

Then use the same naming rules across the team.

CustomerCaseStudy, customer-case-study, case_study_video and testimonial2026 may all describe the same campaign to a human.

Analytics will not necessarily treat them as the same thing.

Also keep personal information out of campaign URLs. You do not need a customer’s name or email address inside the link simply to identify where the click came from.

Test the journey before publishing the campaign

This is boring.

It is also where a surprising amount of measurement fails.

Click the real campaign link.

Does it reach the correct page?

Do the tracking parameters survive any redirect?

Does the video actually play?

Does the enquiry form work?

Does the enquiry reach the right inbox or CRM?

Does the thank-you step record whatever event you intend to report?

If the intended customer journey is:

LinkedIn clip → case-study page → enquiry form

test that exact journey on a real phone and desktop browser before spending money sending people through it.

There is little value in an elaborate ROI report if the campaign link was broken for the first week.

Use views and watch time to diagnose, not declare victory

Imagine the case-study clip gets 50,000 views but no meaningful enquiries.

That does not automatically mean the video failed.

Perhaps the objective was broad awareness.

But if the stated objective was qualified leads, views are not enough to call it a success.

They are useful because they help locate the problem.

Hardly anybody sees the video

The issue may be distribution.

Making a different edit will not help much if the intended audience never receives it.

People see the post but do not watch

The opening, thumbnail, placement or promise may not give them a reason to start.

People watch but never visit the page

Perhaps the next step is unclear.

Or perhaps the film already gives them everything they needed and there was no reason to click.

People visit the page but do not enquire

Now look at the page, offer, form, pricing, trust signals and audience quality.

Do not automatically blame the video.

Lots of enquiries arrive, but they are poor fits

The video or campaign may be attracting the wrong people — or failing to qualify them clearly enough.

This is why engagement metrics belong in the diagnosis.

They help explain where the journey changes, not how many pounds the film made.

Show the counts alongside the percentages

Small businesses can produce very impressive-looking percentages from tiny numbers.

Last quarter:

2 qualified enquiries.

This quarter:

4 qualified enquiries.

That is a 100% increase.

It is also two additional enquiries.

Both facts should appear in the report.

The same applies to conversion rates, applications and completed sales.

Percentages make comparison easier.

Counts stop the comparison becoming misleading.

For a niche B2B campaign where one large sale can completely transform the numbers, this matters even more.

Record what else changed

Suppose enquiries rise after the video launches.

During the same month you also:

• reduced the price;
• rebuilt the landing page;
• doubled paid-media spend;
• hired a new salesperson;
• attended a major industry event;
• and entered the busiest seasonal period of the year.

The video may have helped enormously.

You still cannot sensibly give it all the credit.

Keep a simple campaign change log.

It does not need special software.

Record meaningful changes that could affect the same outcome you are measuring.

Then your report can say:

“Qualified enquiries increased from 14 to 21 during the campaign. The video launched during the same period as a revised landing page and a 25% increase in paid-media spend, so the increase cannot be attributed to video alone.”

That is a much more useful statement than manufacturing false certainty.

Pipeline is not revenue

A £100,000 sales opportunity is exciting.

It is not £100,000 of return.

Keep these figures separate:

Qualified enquiries
People or organisations that genuinely fit the service.

Opportunities
Potential deals being actively considered.

Pipeline value
The possible future value attached to those opportunities.

Completed revenue
Work actually won.

Contribution or profit measure
The value remaining after whatever costs your agreed calculation includes.

Probability-weighted pipeline can be useful for forecasting, but it is still an estimate.

Do not quietly add open opportunities to completed sales because the ROI slide looks better that way.

Some useful videos do not need a sales ROI calculation

Trying to convert every video into pounds can become just as misleading as measuring everything by views.

Recruitment

A recruitment film might be judged against:

• suitable completed applications;
• applicants withdrawing after better understanding the role;
• acceptance rate;
• repeated candidate questions;
• recruitment-team time;
• or retention where the organisation has enough evidence and time to assess it.

Fewer applications could even be positive if the people who decide not to apply were poor fits.

Training

A training video could be measured through:

• completion;
• assessment results;
• error or rework rates;
• manager time spent repeating the process;
• or employee confidence where it is measured appropriately.

If a twenty-minute video saves an experienced manager from repeating the same forty-minute introduction 60 times per year, there is a potential time value.

Calculate it transparently.

Do not invent a pound value for “engagement” simply because the finance team asked for ROI.

Customer support

A clear demonstration might reduce repeated support questions.

Measure the relevant ticket volume and the quality of the outcomes before and after, while accounting for other changes to the product and support process.

Events and internal communications

A recording might exist because people physically could not attend live.

An internal update may exist because everybody needs to hear the same explanation.

The return may be operational consistency, access or avoided repetition rather than direct sales.

Report the outcome honestly instead of forcing the video into an inappropriate revenue model.

A reusable film changes the cost calculation

Suppose the £4,000 customer film is used for:

• the original three-month campaign;
• sales meetings for the next year;
• an exhibition;
• two later email campaigns;
• the customer’s own partner page;
• and several social extracts.

What is the production cost of the second campaign?

You need an accounting rule.

You could charge the complete production cost to the first campaign and treat later reuse as additional value.

Or you could allocate the production cost across an agreed useful period.

Either can be reasonable.

What matters is using a method consistently and making it visible.

Do not allocate all the cost to one campaign when performance is strong and suddenly spread it across three years when performance looks disappointing.

Our guide to how much a business should spend on video marketing looks at the wider campaign budget, including distribution and measurement rather than production alone.

Your tracking will always have blind spots

Not everybody can or should be followed perfectly across the internet.

People change devices.

They decline optional tracking.

They speak to colleagues offline.

They type the URL directly.

They watch a film in a sales meeting.

They see something on LinkedIn and enquire three weeks later after searching for the company name.

Some journeys will remain incomplete.

Treat missing information as missing information.

Do not treat it as proof that the video had no effect.

Website tracking also needs to be implemented within the applicable privacy rules. The ICO finalised updated guidance on storage and access technologies in April 2026, covering technologies including cookies, tracking pixels, link decoration and similar mechanisms under PECR and, where relevant, data-protection law. (ico.org.uk)

Have the people responsible for your website and data protection review the actual setup.

Do not assume that calling a script “analytics” automatically answers the compliance question.

The monthly report should end with a decision

A report does not become useful because it contains twelve charts.

At the end of the reporting period, somebody should be able to answer:

What are we going to do differently because of this?

Perhaps:

• keep distributing the case study because qualified enquiries are strong;
• stop paying to promote the broad version and test a more specific extract;
• move the video higher on the page;
• create another case study for a different customer type;
• fix the enquiry journey;
• give sales a better way to send the film;
• or stop the campaign because the evidence does not justify more spend.

A compact report might contain:

That is enough if it helps someone make a decision.

Put the measurement plan in the brief

The worst time to decide how a campaign will be measured is after it has finished.

Before production, agree:

• what the video is supposed to help change;
• the existing baseline where one exists;
• where the film will appear;
• the intended next action;
• how campaign routes will be distinguished;
• who owns the reporting;
• and when the result will be reviewed.

Those decisions can affect the video itself.

If the next step is a product demonstration request, the film should support that.

If sales needs a case study to send directly to prospects, the production may need a self-contained version that works outside a public campaign.

If recruitment is the job, asking for “more views” in the production brief solves the wrong problem.

Our guide to writing a video production brief helps put that information together before filming.

And if you are still deciding whether a larger production makes financial sense at all, our article on whether professional video production is worth the investment includes a simpler break-even test.

At The Camera Guys, our corporate video production can be planned around the result the business actually needs rather than a target view count.

Send us the outcome you are trying to change.

The cameras come after that.