Measuring whether the video beat the PDF

Why views versus downloads is meaningless

A view on most platforms is counted after two or three seconds of autoplay. A download requires a deliberate click, and often a form. These are not comparable units, and comparing them makes video look ten times better than it is.

The comparison that means something is completions: how many people consumed the whole argument, in either format.

Metric 1 — completed consumptions

For video, take views multiplied by average percentage viewed, or better, use the platform's count of viewers reaching the end. For the PDF, this is harder but not impossible: a hosted PDF viewer or a tracked download plus a scroll-depth proxy on an HTML version gets you close.

The typical finding for document content is that the video's completion count exceeds the PDF's substantially, while its view count exceeds it enormously. The first number is the real one.

Metric 2 — cost per completion

Divide production time by completions. The PDF's production cost is already sunk, which is what makes the video's marginal cost the relevant figure: you are not comparing writing a report against making a video, you are comparing making a video against doing nothing more with a report you already wrote.

That framing usually makes the video look good, and it is the honest framing rather than a favourable one.

Metric 3 — downstream action

Completions are engagement, not outcome. The question that decides whether to keep doing this is whether the video produced the thing the document was for — a demo request, a citation, a reply, an enrolment.

Attribute carefully. The most common pattern is that the video is discovered first and the document is the conversion surface, which naive last-touch attribution reports as the document doing all the work and the video doing none.

• Tag the link from the video description distinctly.
• Look at assisted conversions, not last touch.
• Track document downloads that originate from the video's page separately.

Metric 4 — reach among non-readers

The strategic reason for making the video is reaching people who would never have opened the PDF. That is measurable: look at the audience overlap between the video's viewers and the document's downloaders.

If they are largely the same people, the video is serving your existing audience differently, which is fine but is not the argument you made for it. If the overlap is small, the video is doing the job.

What a fair test looks like

Publish both, promote both equally, and give it a quarter. Comparing a video you promoted hard against a PDF sitting on a resources page measures your promotion, not the format.

And run it on more than one document. Format effects are smaller than content effects, so a single comparison mostly tells you which document was better.

Frequently Asked Questions