Tremor Video Inc.'s VideoHub analytics business is upping its profile today with two big moves.
It will formally introduce VideoHub eQ score, a patent-pending formula to decide the potential of video ads to get viewers’ attention, a gee-golly undertaking that measures how ads perform and are seen—and using 14 ways to distinguish types of viewers or situations. It gives a quality grade that tells advertisers if the ads are doing what they ought, placed on the right sites or are hitting the target audience.
VideHub is also announcing a partnership with comScore which will begin using VideoHub to measure video ad viewability based on a technology that VideoHub debuted last May. Teaming with comScore gives VideoHub analytics a shiny credibility trophy and no doubt ups its influence in the biz.
“It's a huge validation,”says Anthony Risicato, VideoHub’s general manager, and since comScore numbers are noted constantly, it’s a sure thing VideoHub’s viewability data will get wide exposure too.
The measurement and viewability technology advances, says Tremor Video CEO Bill Day, “is one of the reasons we started the company.”
The announcements coincide with chief revenue officer Randy Kilgore’s ascension as new chairman of the Interactive Advertising Bureau, announced Sunday at the IAB’s annual leadership meeting in Phoenix. So that are three pieces of news right there, and it’s just Monday..
The eQ score is a major endeavor, and ought to cause some sizeable waves of interest in the online video business. It sizes up video impressions by viewability (percent of the video ad seen on the screen), player size (under 300 pixels to over 800 pixels) and completion rate (how much a viewer cares to watch,.) Ads get a score of 0 to 100, though what’s a good score really depends on what the marketer was hoping to accomplish. (But, says Risicato, if you look at it like scores you got on homework, you’d often be appraising the numbers in a way that makes sense.)
VideoHub player size is measured when the ad begins, and uses four 7.5 second quartile segments to measure viewability and completion rate.
VideoHub provides a formula for scoring each video ad impression which goes like this: eQ=Player Size Value X Average Viewability Score over Duration of Ad. Sorry, publishers, if a viewer splits before before seeing the whole ad the quartile score is a big zero, which brings down the total score.
The formula was in development for about a year, and then went through months of more testing. “We worked with a handful of agency partners who really dove into the analytics pretty deeply and began looking at time of day and video view history and demographics and looking at publishers and placement that were having the best impact,” Risicato relates. "Some of them began to develop way to cobble this together holistically and we worked very closely with them. Our folks began looking at this very carefully.”
Risicato says eQ can look at ads and how they performed by separate characteristics like geography, content category, time of day, day of week, publisher, weather, and even browser type since there are some advertisers who are looking to certain kinds of computer users.
The wealth of data is then sifted through the eQ formula and gets a grade, presumably freeing marketers from what Risicato calls the “tyranny of data.” But, since there’s an awful lot of data involved with eQ scores, too, he certainly means the oppression that results from data that’s up for random interpretations when not teamed with other moving parts. Reminds me of something like an analytic Slinky that ends up navigating the stairs with more and more momentum.
A little surprisingly, some of those data points tend to lose their significance, when you’re counting impressions by bushels of millions. Like, Risicato says, take male-female demos. With huge numbers of viewers, “it doesn’t drive behavior as much as we thought,” he says, except for some very girly-girl (cosmetics?) or manly man (beef jerky?) products. “Men and women are not as different as we think,” Risicato says.
Or as easy to pigeon-hole, as a result. In experiments with some participating brands, eQ applied the analytics to a yogurt that was aiming for urban professional women, but “it turns out they were attracting suburban young men,” Risicato’s data showed. He was about to break this awkward news at a meeting with the company when its own analytics team burst in with reams of paperwork detecting that the brand was attracting busloads of male college athletes consuming yogurt to control their playing weight. That, presumably, gave backbone to eQ's own data and a way not to get kicked out of the conference room.
“We get a lot of interesting outlier information when we look at this data,” Risicato says. “We see a campaign that has two placements, one trying to drive clicks, another one trying to drive brand health, and their frequencies are at the opposite ends of the scale. Maybe it’s a 10 to drive brand health and 1 or 2 for drive clicks and for the most part a marketer hasn’t known that. What impact is frequency really having? They usually use it as a cost-control mechanism, not an effectiveness mechanism. That’s actually one [example] that I spend a fair amount of time looking at because I find that so fascinating.”
As for the comScore alliance that will highlight VideoHub’s viewability analytics, Risicato thinks the data can be a boon for publishers. “I wouldn’t be surprised if
some publishers who have high viewability don’t begin using them and saying to advertisers, ‘Buy here if you want to be seen.’ "
The measurement will be provided according to the working IAB definition of a viewable video ad—50% of the video player in view for at least half the ad. The measurement will occur through and IAB standard VAST/VPAID ad tag. As long as publishers use the tag it can be measured. That will make it of instant use on YouTube, Hulu, Yahoo! and more.