eMarketer has some interesting data up today, pulled from both Limelight Networks and Freewheel, showing that shorter online video ads are watched to completion more often than longer video ads. And that seems totally logical. However, the data also reveals that viewers are much more willing to watch ads if all lengths when the actual root video content the ad is attached to is longer. 

Shorter Ads More Likely To Be Watched

Now, the first part is common sense: of course people are more likely to complete a 15 second ad than they are a 30 second ad… it’s only half as long. Here’s the graph with the actual numbers:

That’s about a 20% difference. Which might… on the surface, encourage companies to start shortening their online video ads. But that would be a mistake, because this study showed that this data is likely to change when you consider the length of the video the ad is placed on.

Viewers Complete More Ads When Root Video Is Longer

Turns out, we don’t just favor shorter ads period. We favor them… in general. But when the video we came to see is longer, we become much more willing to tolerate longer ads. Here’s eMarketer’s chart, using Freewheel’s data, showing how ad completion rates change between 15 and 30 second ad lengths when the root content is longer:

There’s almost no change in our completion rates between 15 and 30 second commercials… when the root video is long. Which leads me to believe that when viewers sit down to watch a piece of online video they know from the start is longer than usual, we end up more willing (probably subconsciously) to accept more advertising with it.

That’s kind of interesting, actually, if you really examine the chart. Because Freewheel’s data seems to suggest that 30 second ads are more watched than 15 second ads across long, medium, and short content… which kind of directly impacts the Limelight Networks data above saying the exact opposite.

Regardless, it’s clear that longer video content encourages more ad completion by viewers.

Ad Placement Makes A Difference

Using more Freewheel data, eMarketer points out that when a video ad runs has a huge impact on whether we watch it to completion or not. Mid-roll ads, for instance, are completed at astronomically high rates. Of course, mid-roll ads (like what Hulu has used for years) are most often not skippable, which means viewers have no choice but to complete the ad if they want to get back to the content.

Pre-roll and post-roll ads don’t fare nearly as well, and even seem to be fading in completion rate over time. Here’s the chart:

Now, to some degree, this makes perfect sense. Of course post-roll videos have the lowest completion rates… they run after the content is over. Many viewers skip out once their program is finished. Same goes for pre-roll… of course it’s lower, we get the “skip” option with these ads more and more every day. But in giving viewers the option to skip ads, publishers are also encouraging lower completion rates.


It’s all a guessing game anyway. Just because I, as a viewer, complete an ad… that doesn’t mean I actually paid attention to it. I could have tuned it out to focus on a dozen other things around me. And just because I fail to complete an ad… doesn’t mean I didn’t retain any of it. Completion rates are a measure of whether an ad was completed… that’s it… they say nothing about whether the ad was effective, which I think is important to point out. Make a crappy ad, and you can have a 100% completion rate that still drives no buzz or sales whatsoever.

Regardless, completion rates are an important stat in the growth of online video spending. Which means we’re likely to start seeing far more mid-roll ads over the coming year, with pre-roll and post-roll continuing to lose effectiveness. Similarly, we will likely see brands getting more savvy with how long their video ads are, and what length of video content those ads are paired with.