Skip to main content

We all know how dominant video has become in the last 18 months across social media. With advances in recording devices (i.e. phones!) and new social network features, video has become super simple and quick to produce and distribute.

Furthermore, and of real interest to publishers and brands, video remains one of the few formats on social which provides a sense of how much time was spent with it… making it even more powerful. However, that’s not to say measuring it isn’t without its challenges.

This post outlines what can make measuring video tricky, and suggests some approaches to simplify.

All videos aren’t created equal…

When we talk about “videos” there are so many variations to bear in mind.

First off, a video can of course be of any duration. Comparing a five-second video with a four-minute piece is unlikely to offer a fair comparison.

Secondly, the way video is presented, and the way people use different social networks shapes how video is consumed.

  • On Twitter and Facebook (for example), with a rapid browsing/ scrolling behaviour, video is dropped into the audience’s newsfeed where it must work hard to get attention.
  • Snapchat and Instagram (ads) pop up between other Stories you’re looking at in a more disruptive style.
  • On the grandfather of video, YouTube, content is typically sought out by the viewer, with episodic content of 10-20 minutes far more prominent.

(There are of course many other formats, and Facebook noticeably is now getting into the longer form space with its Watch tab feature).  

Finally, and most importantly when it comes to assessing performance we must consider different objectives. Was the video trying to ultimately get people to click through to a website? Was it to encourage people to tell their friends about it and, by extension, the brand?

There’s plenty to measure

Speaking of measuring performance, there certainly isn’t a lack of metrics available. Some of the most prominent include:

  • Views: to understand how many people actually saw a video (see note below).
  • Watch time: total amount of time all viewers spent watching a video.
  • Percent completion: what percentage of a video the average person watched.
  • Views with sound on: how much the video drew in the viewer and encouraged them to be more immersed in the creative.

Screen Shot 2017-10-19 at 12.26.58.png

However, this is where more of the nuance comes in. A couple of key things to consider are:

  • Different social networks define “a view” in different ways, and there’s a whole debate as to what fairly constitutes a view. Should 3 seconds of a video that auto-plays without sound be deemed as a view? (discuss!)
  • Facebook provides some measures for organic video (e.g. 3 sec, 10 sec, 95% completion), but then further ones only for paid videos, such as 25% or 50% completion. Which makes it more difficult to compare apples with apples.

When things can start to get a little messy

Having all of these options means you can get right into the detail of what happened around an individual video. However, when you start trying to take more of a macro view, the amount of variables can become problematic. We’ve already discussed different objectives, different social networks, paid vs organic, video length and a host of different metrics… and we’ve not even touched on different audience groups!

To have a chance of seeing the wood for the trees and understanding how well different types of video are performing, some sort of barometer or benchmark is required.

Being brutal about variations

There’s no one-size fits all approach, but here are a few ideas on simplifying these variations from a reporting perspective…

  • Keep social networks separate: trying to group performance across different social networks is very problematic, it’s best not to combine them as they are just too different.
  •  Video length: can you group videos into perhaps two groups – short (under 15 seconds) and long (over 15 seconds)? The first delivering short sharp bursts of information versus longer form “stories”, such as a how-to guide or recipe.
  • Video metric: will something like 50% video completion work as a common measure for all videos? If someone gets through half of a short video, they’ve arguably taken on board much of the information it’s trying to convey. And if someone watches at least 50% of a long video, it shows they’ve decided to spend a chunk of time (by social network browsing standards) with your brand.

The above might be considered too simplistic. Or you might feel that different completion percentages are appropriate for longer/shorter pieces of content. But hopefully what it can do is free you from drowning in data by trying to assess too many variables… and subsequently extracting no insights.

Finding patterns

Beyond the above approach, you’ll of course want to run more in-depth analysis around the way different video styles are performing (e.g. person speaking to camera, vs hands in shot demonstration). By filtering on the style of video content first and then analysing performance across your chosen metrics should help you extract some actionable insights.

A couple of further considerations

  1. Another approach you might be interested in is assessing the ratio of paid to organic video views… this can be an effective way of further illustrating the impact of different creative.
  2. A whole other conversation exists around different types of media-buy for video against different “viewability standards”, such as a 2 second view, completed views or just impressions. It’s certainly one to keep an eye on as the social networks and media agencies grapple with it.

In summary

  • Video is great because you get a sense of time spent with the brand.
  • However, there are many nuances around measuring it.
  • The most important thing is having a clear understanding of the video’s objective.
  • Simplified reporting is needed to spot patterns.


Is video critical to your content marketing plan? Which metrics do you use for measuring it?

One Comment

Leave a Reply