Size matters, but not in the ways you might think for video advertising.
First, the fast expansion of video inventory is not driving down pricing. Both video ad volumes and prices are increasing pretty much in tandem through 2010, reports FreeWheel in its quarterly Video Monetization Report. Indexed against the full year of video pre-roll ads served through its system, the company found that video ad views grew 20 points through 2010 while CPMs were up 19 points on an index basis. FreeWheel data included 23 billion video views in 2010 and 13 billion ad views. The number of pre-roll ads served in the last quarter of the year was up 97% over the previous quarter.
Second, ad length does not affect ad completion rates as much as overall content length does. At the same time pre-rolls were proliferating, longer video sessions and smarter placement strategies found that the mid-roll volume more than doubled in the quarter (+216%). In fact, the mid-roll enjoys the best ad completion rate (91%) of all formats even as pre-rolls represent 87% of the market. Nestled within a longer online viewing experience, the mid-roll gets a completion rate of 81% when within 20+ minutes of programming and a 62% completion rate when interrupting mid-length video of 5 to 20 minutes. Shorter content lengths equal lower ad completion rates.
Generally, completion rates in pre-roll and mid-roll units are affected more by the length of the surrounding video than the length of the ad itself. Apparently if viewers feel they are getting value from the content, then the difference between sitting through a :15 or a :30 is negligible. In long form content, :15s had a 82% completion rate compared to 86% for longer form. Clearly people are acknowledging the nature of the ads-for-content value exchange, especially when the media experience is prolonged. Fifteen-second ads still make up 58% of the market in FreeWheel's ad serving world, and remarkably only 6% are shorter than that.While FreeWheel's data is revealing about user behavior among those who are in the video experience, the effect of longer pre-rolls on user willingness to initiate a video at all on sites where the ad-to-content ratio is out of whack is not as clear. Numerous major news site brands deter me from starting video clips because of their tendency to put :30s in front of one-minute news clips. The value exchange is so wrong so regularly that I hesitate from engaging their video content at all. The effect of ad length on video usage patterns can be felt, I suspect, in ways that basic video metrics can't measure.