How to Measure the ROI of Video Marketing

Choose the outcome before choosing the dashboard

Start with the reason the film exists. A product demonstration might help buyers qualify themselves. A customer story might support sales meetings. A recruitment film might help applicants understand a role. These should not all be judged against the same number of views.

Write down the primary outcome, the measurement period and the person responsible for reporting it. Define terms too. Does an enquiry mean every form submission, or a genuine potential customer who fits the service? Does an application mean a started form or a completed suitable application?

Keep the reporting chain short enough to explain: relevant audience reached, meaningful engagement, next action, business outcome. The middle steps help diagnose a problem. They are not interchangeable with the final result.

Use a financial calculation with a clear boundary

A practical campaign calculation is:

ROI (%) = (attributable contribution before campaign cost − campaign cost) ÷ campaign cost × 100.

Here, contribution means revenue remaining after the relevant costs of fulfilling that work, before subtracting the campaign cost. Agree the accounting basis with your finance team and use it consistently. Do not subtract the same cost twice.

For an illustrative example, suppose a campaign costs £6,000 in total and is credited with £30,000 of sales. If those sales provide £12,000 of contribution before marketing cost, the calculation is (£12,000 − £6,000) ÷ £6,000 × 100 = 100% ROI.

Dividing £30,000 revenue by £6,000 campaign cost gives a revenue-to-cost ratio of 5:1. It is not the same calculation. Neither figure proves that every credited sale was caused by the video; that attribution question still needs answering.

Include the complete cost of the campaign

Count planning, filming, editing, versions, captions, media spend, hosting or page work specifically required for the campaign, and the internal time included in your chosen cost basis. Record what is excluded.

A film reused over several campaigns needs a stated allocation method. You might charge production to the initial campaign and show later reuse separately, or allocate a share across a defined period. Choose one approach and make it visible. Changing the method to improve a disappointing result makes comparisons meaningless.

The video marketing budget guide helps identify the cost lines. Our production investment guide covers the earlier decision about whether professional support is proportionate.

Record a baseline and the other changes happening

Before launch, record the relevant existing performance: qualified enquiries, product-page conversion, sales-cycle questions or another outcome suited to the brief. Use a period long enough to avoid mistaking one unusual week for normal behaviour.

Keep a change log. A new price, seasonal demand, a trade show, a larger advertising budget or a redesigned enquiry form can affect the same outcome. If enquiries increase after all five change, the video cannot honestly receive all the credit.

For small B2B audiences, a handful of deals can dominate the result. Show the counts alongside percentages. Moving from one sale to two is a 100% increase, but it is still only one additional sale and may not support a confident forecast.

Give each distribution route an identifiable link

Use consistent campaign tagging on links from email, social posts and other external placements where appropriate. Google’s campaign URL guidance explains UTM parameters for identifying the source, medium and campaign that referred traffic.

For example, use a stable campaign name and distinguish the email extract from the paid social edit. Keep a shared naming sheet so one team does not use five spellings of the same campaign. Do not put names, email addresses or other personal information into campaign URLs.

Test the whole journey before release. Follow the real campaign link, load the page, play the video and complete the intended test action. Confirm that redirects preserve the useful campaign information and that the enquiry reaches the correct team. Remove test records from business reporting through the normal agreed process.

Connect enquiries to sales outcomes

A form submission is the start of a relationship, not automatically revenue. Where your systems support it appropriately, carry the campaign reference into the customer record and follow the enquiry through qualification, opportunity and completed work.

Ask the sales team what they actually used. A prospect may arrive through search and later watch a case study sent by a salesperson. A last-click website report may never show that contribution. A simple recorded note can provide useful context without pretending to be a controlled experiment.

Keep pipeline value separate from completed revenue. A £50,000 opportunity is not £50,000 earned. If you report probability-weighted pipeline, label the probabilities as estimates and revisit them when opportunities close.

Attribution assigns credit; it does not settle causation

Google Analytics attribution documentation explains how its models assign credit across interactions. The result depends on the model, available data and settings. It is a reporting view, not a complete record of everything that influenced a person.

Platform reports can overlap. An advertising platform may credit a sale after someone saw an advert while another credits a later click. Adding every platform’s claimed conversions together can count the same customer more than once.

Where practical, use a suitable comparison or controlled test: similar audiences, a pre-agreed outcome and enough observations to support an interpretation. Keep other major variables stable. With low volumes, report directional evidence and the limits rather than inventing statistical certainty.

Use engagement to decide what to improve

Look for the point where the intended journey breaks. Few relevant people seeing the film suggests a distribution problem. People seeing the player but not pressing play may suggest weak placement or an unclear promise. Early exits may suggest the opening takes too long to reach the useful point.

People watching but not enquiring may already have received all they need, or they may face a weak offer or difficult next step. Do not assume every drop-off requires a new edit. Review the film in the page and campaign context.

Compare like with like. Platforms define views and engagement differently. A short autoplay social clip and a deliberately opened technical demonstration do not share the same viewing conditions.

Respect the limits of tracking

Some journeys will be unobserved because of consent choices, device changes, offline conversations or technical limitations. Treat missing data as missing data, not proof of no effect.

Have the website team check the actual measurement setup against current ICO guidance on storage and access technologies. Requirements and exceptions depend on purpose and implementation. A tool being described as analytics does not by itself resolve the privacy question.

Collect only what the measurement task needs. Reporting aggregate campaign outcomes is often more useful than accumulating detailed individual viewing histories that nobody has a justified plan to use.

A useful monthly report fits around decisions

• Delivery: what was published, where and when.
• Audience: who was reached and whether that matches the brief.
• Response: meaningful engagement, qualified actions and actual counts.
• Commercial outcome: completed contribution, with pipeline shown separately.
• Cost and attribution: included costs, method and uncertainty.
• Next decision: continue, change distribution, adjust the message or stop.

For a training or support film, use appropriate operational measures instead. Time saved needs a defensible baseline; satisfaction and understanding can be reported without forcing them into a fictional pound value.

Common video ROI questions

Put measurement in the production brief

Tell the production team what the film needs to help achieve and where it will be used. That affects the interview questions, opening, versions and next step. Use the briefing guide, or discuss an outcome-led production with The Camera Guys.