Measurement
Part of A method for measuring and reporting short form video performance that avoids vanity metrics
A short form video reporting dashboard that survives a board review
Build a short form video reporting dashboard that ties platform metrics to business outcomes, with a scoring rubric, consent checks and a weekly refresh routine.
What to take away
- Decide first what the dashboard must settlewhether short form video earns more budget, keeps its current share, or stops. Every panel should serve that verdict.
- Report three layers togetherreach, engagement and a business outcome such as qualified leads or sales. A dashboard holding only views cannot answer the budget question.
- Fix the definitions before you build. Agree what counts as a view, a qualified view and a converted viewer, then lock them for at least a quarter.
- Check your consent and tracking basis before adding any audience panel, because electronic marketing rules shape what you may collect and store.
- Publish on a fixed weekly cadence with a named owner, so the numbers arrive before the meeting rather than during it.
Agree the questions and the data model
Start from the decision, not the chart
Write down the decision the dashboard exists to support. For most teams it is a budget or format call: keep vertical video, expand it, or move spend elsewhere. That single sentence tells you which metrics matter and which are decoration. If a panel would not change the decision, leave it out.
Data dictionary essentials
- Metric defined in platform's own wording
- Date the definition was checked
- Source system named
- Refresh time recorded
- Owner named for each column
Define every metric once
Ambiguity kills dashboards. A view on one platform is not the same event as a view on another. A three-second play is not a completed watch.
Use these house definitions and preserve each platform’s native figure separately. A view is the view count reported by the platform, using its own definition; do not treat counts from different platforms as identical. A qualified view is a reported view with at least three seconds of watch time, where the platform export supports that measure; otherwise mark it unavailable. A converted viewer is someone linked to a named outcome within the stated attribution window under your documented method; report the outcome as attributed, not necessarily caused by the video.
Use native analytics sources for short-form video: TikTok Analytics or TikTok Ads Manager; Instagram Insights or Meta Ads Manager for Reels; and YouTube Studio Analytics for Shorts, with Google Ads for paid YouTube campaigns. Keep organic and paid exports labelled separately. The Bing Ads API overview covers Microsoft Advertising, so exclude it from short-form video source mapping.
Keep a short data dictionary beside the dashboard. Name the source system, the refresh time and the owner for each column. Anyone joining the team should be able to trace a number back to its origin without asking.
Board dashboard template
Use one row per platform and placement, and keep the native source visible so figures can be traced back to an export.
| Platform and placement | Native source | Spend | Impressions | Native views | Qualified views | Average watch time | Completion rate | Outcome | Cost per outcome | Refresh and owner |
|---|---|---|---|---|---|---|---|---|---|---|
| TikTok video | TikTok Analytics or TikTok Ads Manager | |||||||||
| Instagram Reels | Instagram Insights or Meta Ads Manager | |||||||||
| YouTube Shorts | YouTube Studio Analytics or Google Ads for paid campaigns |
At the top, show the reporting period, the budget decision, total spend, the selected outcome and its cost. Do not combine native view counts into a cross-platform total. Add a footnote with the house definitions, attribution window and any unavailable measures; record the data owner and last refresh.
Build the dashboard in three layers
Layer one: reach and engagement
This layer covers impressions, views, average watch time and completion rate. It is the easiest to populate and the easiest to over-weight. Treat it as context for the layers beneath, not as the result itself.
Three dashboard layers
- Layer one: reach and engagement
- Impressions, views, watch time, completion rate
- Layer two: outcomes and cost
- Leads, calls, revenue, cost per outcome
- Layer three: routine and scoring rubric
Segment by format and placement rather than lumping everything together. Vertical video in a feed behaves differently from the same clip in a story or a paid placement. Boosting can extend organic reach on some platforms, and LinkedIn's own guidance on boosting posts sets out how promoted reach is handled there.
Layer two: outcomes and cost
Tie the same period to a business measure. Depending on your model that might be qualified leads, booked calls, or revenue attributed within a stated window. Show cost per outcome beside volume, so a rise in leads with a larger rise in spend is visible rather than hidden.
Use labelled illustrative figures when you model targets. For example, a team spending £2,000 a month to generate 40 qualified leads is paying £50 per lead, and can test what happens if spend rises 20 per cent. Our guide to short form video measurement and reporting explains how these layers fit a wider reporting structure.
Set the routine and the scoring rubric
Consent, tracking and the legal basis
Before you add audience or remarketing panels, document the lawful basis for processing personal data under the UK GDPR and follow the Information Commissioner's Office (ICO) guidance. PECR may also require consent for activities such as storing or accessing information on a device or sending electronic marketing; its 2019 amendment regulations show how that framework is updated over time. Record where consent was captured where applicable, how it can be withdrawn, and how the data will be used.
Score the dashboard each quarter
Run a short review against fixed criteria. Score honestly, then fix the lowest line before adding anything new.
| Criterion | What good looks like | Score (1-5) |
|---|---|---|
| Decision fit | Every panel supports a named decision | |
| Definitions | Each metric defined and dated | |
| Outcome link | At least one business measure present | |
| Consent basis | Lawful basis recorded and reviewable | |
| Refresh routine | Weekly update, named owner | |
| Traceability | Any figure traceable to source |
Anything scoring 2 or below becomes the next sprint's work. Compare your own numbers against sector ranges in our short form video benchmark research before you conclude a metric is strong or weak.
Common questions
How often should the dashboard refresh?
Weekly suits most marketing teams, with a monthly roll-up for budget owners. Daily refreshes invite reactions to noise, and quarterly cycles arrive too late to change a campaign already running.
What belongs on the first screen?
Three things: the outcome measure, its cost, and the trend against the previous period. Reach and engagement sit below, as supporting evidence for the outcome figure.
Should organic and paid video share one dashboard?
Yes, but keep them in separate columns. Blending them hides whether paid spend is lifting results or simply replacing organic reach you already had.
How do we handle platforms that report differently?
Define a house metric for each concept and map platform figures onto it. Document the mapping, then state it in a footnote so readers know comparisons are approximate.
