How to structure a video marketing reporting dashboard

What to take away

What to measure before you build

Agree the metric set with the client

A dashboard is only as good as its definitions. Before any build, write down what each metric means, which platform produces it, and who owns it. Reach and view-through rate mean different things across YouTube, LinkedIn and a website player. If two teams disagree on a definition, neither number is trusted.

Start from the client's commercial question. A retailer wants assisted conversions. A public body tracking behaviour change may want recall and sentiment. Whichever you choose, split the dashboard into two layers — what the platform reports and what the business records — before you commit to a schema; the measurement and reporting overview shows how that split fits a wider programme.

Write the metric dictionary once and version it, so a platform change records a new definition and the old one stays visible.

Most agency stacks are built on one of four tools: Looker Studio, which connects directly to YouTube and Google Analytics 4; Power BI; Tableau; or a connector such as Supermetrics that pulls platform data into a spreadsheet or warehouse first. The tool matters less than the definitions inside it, so agree the dictionary before you open the tool.

Platform definitions differ, so the same campaign produces three different totals: YouTube, LinkedIn and a website player each apply their own threshold before a play counts as a view, and each reports retention in its own way. State the definition on the chart, and never place a platform view beside a site-player view without a label.

Separate platform metrics from business outcomes

Platform dashboards report what the platform can see, rarely what happened after the click. Join platform data to analytics, CRM or sales systems in a second layer that answers revenue and enquiry questions rather than impressions.

Where the client gives no analytics access, agree a proxy measure and label it clearly. A labelled proxy is honest. An unlabelled one is a future dispute.

Label each layer with its refresh frequency, so nobody reads a weekly figure as live data.

A workable structure is three tabs. The first, headline, carries the agreed metrics with their definitions on the page. The second, channels, breaks the same metrics down by YouTube, LinkedIn and the site player. The third, sources, records the refresh frequency, the owner and the lawful basis for each field.

What the rules require

Handle personal data lawfully

A dashboard that links viewing behaviour to an identifiable person is processing personal data. Under the UK GDPR guidance and resources from the ICO, you need a lawful basis, a retention period and a route for access requests. In practice two bases cover most video reporting: consent, where tracking rests on cookies or email matching, and legitimate interests, where you record a balancing test. Set the basis per data source rather than once for the whole dashboard. That covers email-matched audiences, logged-in video portals and lead forms.

Keep personal data out of the reporting layer where you can, aggregate before visualising, and store the basis with the metric definitions.

Set a retention period in months and let the dashboard aggregate or delete automatically, rather than relying on memory.

Access requests are the part clients forget. The ICO's right of access guidance sets out what a response must contain, and the process needs an owner and a documented route: who receives the request, who searches which system, and how the answer is recorded. The dashboard also belongs in your record of processing activities, the register UK organisations are expected to keep.

Keep ad claims honest

If the dashboard feeds a public claim such as "watched by two million people", the Consumer Protection from Unfair Trading Regulations 2008 set the standard for what counts as misleading. A figure that counts a three-second autoplay as a view may be accurate and still misleading in context. Put the definition next to the number.

How to present it

Build the before and after view

The clearest dashboards show a baseline and a change, not a single figure. Capture the pre-campaign position before work starts, then show the movement. The table below is an illustrative example for a team spending £8,000 a month on video.

Agree the baseline window before work starts, because a baseline chosen afterwards is easy to dispute.

Add impact and well-being measures

Reach and completion rates miss how a campaign changed a community or a workforce. The ONS work on measuring progress and well-being gives a framework for social impact measures alongside commercial ones, built on four personal well-being measures: life satisfaction, whether the things you do in life are worthwhile, happiness and anxiety. Those four are enough structure for a repeatable questionnaire.

Add a short qualitative panel, such as a short survey, and report it beside the quantitative layer. Keep it repeatable: a consistent quarterly measure beats a one-off study.

Report the sample size beside any survey result, and do not compare waves that used different questions.

Common questions

How many metrics should a dashboard show?

Fewer than most teams expect. Five to eight headline metrics — reach, view-through rate, completion rate, cost per view, assisted conversions and recall — each with a definition and an owner, hold attention. A long list of platform exports usually signals that nobody agreed what matters.

Do we need a lawyer to sign off the dashboard?

Not usually. You need a written record of the lawful basis, retention and access process. Most agencies can produce that with a data protection lead. The disputes that reach lawyers usually start with the same two mistakes, an undefined metric and a platform view shown beside a site-player view.

Can we reuse the same dashboard for every client?

The structure can be reused. The metric set cannot. A template saves build time, but definitions and sources must be agreed per client or the numbers will not survive scrutiny.

How often should the dashboard be reviewed?

Quarterly is a reasonable default, and after any platform change, new data source or public claim. A dashboard that is never reviewed slowly stops matching the campaign it describes.