Video Streams vs. Page Views

July 6, 2009

As I mentioned in my last post, Video Tracking, video is starting to become one of the more popular ways that sites are communicating and/or sharing information with their users. With the increased focus on videos, that begs the question, are the video plays hurting page views? This is a very interesting question. I am going to share some of the experiences that I have had of the last couple of years with the sites that I have worked since we have had more of a focus on video plays. The media sites that I work on also have tv channels that go along with the sites. Whenever possible, we publish full-episode videos to get users to come to the site to watch the episode that they missed on tv. Though there is also content about the show on the site, our theory that we had was that was that the video plays were hurting page views. Did this theory hold true?

As I mentioned before, we had content on the site to supplement the videos, but were users viewing that additional content. Depending on the site and the user’s reason for coming to the site, some of the users would view the additional content. But how much of it did they view. We have put video all over the site to ensure that user’s who want to watch videos; will be given the opportunity to do so. To help me break this out and take a look at what the video users were doing on the site, I used the Omniture Discover onDemand product. Using Discover onDemand, gave me the ability to segment the users who played video. How I defined video users, were users who actually started watching the content. With most media sites, the videos have a pre-roll ad that plays before the content. Also, a lot of sites have turned on auto-play on the video so the users are going to be exposed to the pre-roll right away. I wanted to make sure I took that into consideration when I built my segment. An argument could also be made that I could have segmented users who completed 10% of the video, but for this segment I went with when users began the content.

The first thing that I looked at what how many total page views did the video users consume. I was very surprised when I saw that video users contributed to 25% of the total page views. I was also equally as surprised when I saw that video users contributed to only 12% of the visits. Though when we looked at page views per visit, video visits viewed 2.5x more pages per visit than non-video users. This was a very basic analysis and we did not take into consideration the purpose for coming to the site or the traffic source from which the video user came to the site. We looked at it from a high level perspective. But even from a high level perspective, these results were very surprising. The next thing we looked at as the time spent for video users vs. non-video users. These results were not all that surprising. For video users, the majority of the users spent 10 to 30 minutes on the site, followed by 1 to 5 minutes, and then 5 to 10 minutes. For the non-video users segment, the majority of the users spent 0 seconds on the site (bounces), followed by 1 to 5 minutes, and then less than 1 minute spent on the site. This really goes to show the each segment really does have a different purpose for visiting the site and that will be one of the next things that we will look at.

One of the final things that we took a look at for this analysis, were how many visits played only one video, two videos, three videos, or four or more videos during a visit. Before I took a look at those results, I first wanted to see the average video plays per visit. I saw that video users watched a little more than 2 videos per visit. I based this calculation on video users segment only, so I could get a more accurate number. The results of this analysis were a little surprising, but not too much with 62% of the video visits watched on one video, 15% watched two videos, 8% watched three videos, and 13% watch four or more videos. We were a little surprised with how the percentage that only watched one video. Taking a look at this analysis, we know that a majority of the users watch only one video, but they typically spend a little longer on the site and they will view more pages per visit than non-video users. The answer to the question does an increased focus and an increase in the number of video plays hurt page views. We found the answer to be NO, an increased focus on videos and an increase in the number of video plays do not hurt page views as the video users turned 2.5x more page views than non-video users. But the video users are still a small percentage of visits. So now the question becomes how we can get more visits watching video and viewing pages during the same visit. One of the other things you might want to consider if breaking down this information by traffic source. That is on my list of projects to get done.

Up next: Tracking Marketing Campaigns (paid search, newsletters, twitter, tinyurls, etc)

Advertisements

4 Responses to “Video Streams vs. Page Views”


  1. […] As I mentioned in my last post, Video Tracking, video is starting to become one of the more popular ways that sites are communicating and/or sharing information with their users. With the increased focus on videos, that begs the question, are the video plays hurting page views? This is a very interesting question. I am going to share some of the experiences that I have had of the last couple of years with the sites that I have worked since we have had more of a focus on video plays. The media sites that I work on also have tv channels that go along with the sites. Whenever possible, we publish full-episode videos to get users to come to the site to watch the episode that they missed on tv. Though there is also content about the show on the site, our theory that we had was that was that the video plays were hurting page views. Did this theory hold true?.…….[read entire story] […]


  2. […] This post was Twitted by dericloh […]

  3. web analytics Says:

    Very interesting post. I will have to review how I am currently using this type of information on my web analytics solution.


  4. Nice Article. I’m looking forward to reading about the twitter work you mention as your next article. I’ve done a bit of Twitter into Omniture work, but can’t help thinking there is much more to do.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: