How to Tell Whether Your Video Actually Worked

A film goes live. A fortnight later someone asks how it performed, and the answer is a view count.

The view count is almost always the wrong answer, and not because it's a bad number. It's because it answers a question nobody asked. Views tell you how many times a video started playing. They say nothing about whether it was watched, understood, believed, remembered, or acted on, which is the entire territory the film was commissioned to operate in.

The deeper problem is that measurement is usually considered after the fact. By then the only available metrics are whatever the platform happens to collect, which is why so many post-campaign reviews end up discussing numbers that nobody would have chosen in advance.

Decide What Would Count as Success Before You Commission

The single most useful measurement discipline costs nothing and happens before production - write down, in one sentence, what would have to change for this film to have been worth making.

Not "increase awareness". Something specific enough to be wrong. Prospects stop asking whether we've worked at this scale before. Candidates arrive at interview already understanding how we work. The sales team stops needing forty minutes to explain what the product does.

Claims like these are testable, often without analytics at all. You can ask the sales team. You can ask the next five candidates. The answer may be qualitative, but it's direct evidence about the thing you actually cared about, which is more than a dashboard usually provides.

Metrics Worth Paying Attention To

Completion and retention

What proportion of people who started watching reached the end, and where the rest dropped out. Far more informative than views. A retention curve showing a cliff at eleven seconds tells you something actionable about the opening; a view count tells you nothing you can use.

What happened next

Whether people who watched went on to do something - visited another page, made an enquiry, stayed on the site longer. The behaviour immediately following the watch is usually a better signal of impact than anything measured during it.

Use by the people who need it

Whether your sales team, recruiters, or partners actually send the film to people. Internal adoption is an underrated and highly reliable indicator - content that genuinely helps gets used without being chased, and content that doesn't sits in a folder regardless of how it performed publicly.

Changes in the conversation

Whether the questions you get asked have shifted. If a film was made to address a specific objection and that objection stops coming up as often, that's the outcome - even if no dashboard records it. Worth asking the people who have those conversations daily.

Qualitative response

What people say, unprompted, about the film. A small number of specific, genuine responses often tells you more about whether something landed than a large volume of passive engagement.

Metrics That Mislead

• Raw views. Inflated by autoplay, counted differently on every platform, and easily bought. A number that goes up reliably regardless of quality is not a measure of quality.
• Likes and reactions. Cheap to give, weakly correlated with anything commercial. Often reflect the distribution more than the content.
• Average watch time in isolation. Useful only alongside the retention curve. The same average can describe everyone watching half, or half watching all of it, which are very different outcomes requiring different responses.
• Any single figure without a comparison. A number means nothing without something to compare it against: your own previous content, a period before launch, or an audience that didn't see it.

A Practical Review, Three Months On

A short review a few months after launch is worth more than an immediate report, because the early numbers mostly reflect the launch push rather than the content.

Five questions worth asking:

• Did the thing we said would change actually change?
• Where in the film do people stop watching, and what's happening there?
• Is anyone internally using this without being asked to?
• What did people say about it, specifically?
• If we made another one, what would we do differently?

That last question is the one most worth asking and the one most often skipped. Most organisations commission video repeatedly and learn very little between projects, because nobody writes down what they'd change while it's still fresh.

Happy to talk it through. Get in touch.

‍