• by Steve Smith , Staff Writer @popeyesm, May 7, 2010

A curious bit of data emerged from yesterday's study of Q1 2010 video metrics from Brightcove and TubeMogul. Twitter referrals to videos on every major category of destination resulted in longer viewing times than any other traffic source. A Twitter referral to a music video averaged a 2:33 viewing time compared to 2:01 of time spent by people coming from Google. Tweets drove viewing sessions of 1:52 on broadcast locations, but traffic coming from Facebook, Bing and Google were all in the 1:37 to 1:38 range. The exception to this rule was Tweets landing on newspaper sites, where Yahoo customers viewed one second longer than Twitter refers.

Which is not to say that Twitter is the biggest source of traffic. It is simply the source that seems to bear the most engaged viewer ... if you can get them there from the micro-blog. In fact, 51.75% of video traffic is direct to the destination, while 38.92% is from Google searches, 5.58% from Yahoo, 2.29% from Bing and .4% from Facebook. As a share of referrals, Twitter isn't even showing up in the report, but TubeMogul tells me that it is the sixth most common referral (.19% of traffic), which makes it a less active referrer of video views than Drudge Report.

Still, it is an interesting exercise to ponder why a Twitter referral would tend to stay more engaged. Clearly the social affinity aspect is at play, because Facebook refers, while not quite as high, are also stronger in engagement than most other sources. After all, a refer from a friend compels a viewer to watch longer, if only to "get" what the friend wanted you to see in the video or just because your social graph can target your tastes even more effectively than a straight search. Or, perhaps social referrals filter out the bad hits more efficiently than a search result. And bear in mind that the sample here is of Brightcove customers only, not the entire Web universe. How that skews the results is anyone's guess.

When asking 100 media customers how they are monetizing video, 50% says that in-stream ads are the dominant form, with 47% using in-page and 37% using sponsorship. The last category is the one to watch. Fifty percent of respondents say they will be looking to add sponsorship to their mix in the future. And the sponsorships can't come fast enough, apparently. Despite the explosion of inventory and buyer interest this past year, only 18% of the media companies considered their online video profitable, while another 10% called it break-even. Nearly 30% of the sample's online video operations were deemed profitable or break-even.

And, yes, this will be the year of mobile, even for video. Only 11% of companies said they currently distribute ad-supported video to mobile platforms, another 54% say they will in the next year.

Yet another platform on which not to be profitable.

The full report is available at Brightcove.