If you read the Reel Weekender thoroughly, like the good readers I know you all are, you'll know that I linked over to Pace Lattin's site, Industrypace.com as he's seemingly being threatened due to his reporting on video ad fraud. Well hot on the heels of that, and quite timely for me, comes new research from our friends at TubeMogul talking about, wait for it... Video Ad Fraud (sort of)!
What have they got to say about it? It sucks. OK, that's not really what they said, but we all know they want to right? Even Jim Louderback, CEO of (my beloved) Revision 3 (love Tekzilla, HD Nation, even iFanBoy has its moments and now BEN HECK!) chimed in on it. Yes, I often go off in tangents, but I put them in brackets, so it's OK right?
Seriously, Mr. Louderback had this to say, which I very much admire:
...For a long time I've railed against the industry practice of counting a play-start -- and particularly an automatically initiated play start, as a video view.
And recently I've come to find the off-site -- or embedded autoplay -- as a particularly heinous malodorous video view. Why? Because many embedded autoplay videos often get streamed to pages that have zero content relationship to the video served, and often play below the fold with audio turned off...
Absofrigginletely! I personally hate autoplay and recently, when I was asked by an ad network to do it "in order to show far more impressions," severed our relationship. Haha suckers, now you get nothing from me!
So Jim (I can call you Jim, right?) was actually writing at Advertising Age and was talking about the recent TubeMogul research into... video ad fraud. Well really the research is into what constitutes a video view.
Fraud In Online Video Getting Worse
TubeMogul came to the very unfortunate conclusion via their research the video (ad and viewing) fraud is, in fact, getting worse! So much for the industry righting itself and getting going in a proper direction on this one. You should be ashamed of yourselves.
Here's the deal, two years ago they did some research into what counts as a view. Now they've done an update and we can compare the numbers. They checked out what constitutes a view on eight sites - blip.tv, DailyMotion, MetaCafe, MySpace, Viddler, Vimeo, Yahoo! Video and YouTube. Over the course of five days they did some tests to determine what was considered a full view, half view, refresh, embed and embed autoplay and which of those were counted. They then used videos, with data such that it would be almost impossible to find it through a search etc. to track views on them and keep others from finding the video accidentally.
What Is A Video View And Who's Counting Views Correctly?
The best of the best was blip.tv who didn't count things like a page refresh as another view and limited the number of views counted to one per IP per session on everything else, from the report:
Still one of the most stringent sites for counting views, blip.tv limits view counts to one per IP address per session. Views for videos embedded into another site (i.e.: Embed (>1/2 view) and Autoplay) were also limited to one per session. If videos were watched again after leaving the site and some time had elapsed, it would count as an additional view.
Well done blip! But you've got company as Vimeo also scored quite high in the report in the 'most stringent' category. MetaCafe was in sole possession of the 'stringent' level of view counting and everyone else, but to the chagrin of us all, came in as less stringent.
What does less stringent really mean? It means that multiple viewings from a single IP, even if consecutive or induced by a page refresh, were counted as views on all of the other sites. It also means all embedded views are counted and, with the exception of YouTube, embedded autoplays are also counted. YouTube didn't count them at all (since 2008), blip.tv limited it to one per IP per session. Everyone else counted embedded autoplays as full views. That's 75% of the sites.
In fact, some of the sites have loosened restrictions. MetaCafe, which was quite stringent in the last research, has fallen quite far. DailyMotion has started including refreshes and embed plays.
|Most Stringent Sites for Counting Views|
|Yahoo! Video||less stringent|
For shame! Discounting embedded autoplays should be industry standard and yet, nearly the complete opposite is the reality. That's just, poor form. Not only do they skew the viewing results of all users and videos on the sites, they also work to defraud advertisers whose ads show pre-roll and even mid and post-roll. There's no guarantee that the viewer is even looking at the video, nor even that the video is actually on the active portion of the screen. I have, in my various wanderings around the web, even seen an autoplay video ad a full 1000 pixels down the side of a page, effectively completely hidden from view, with no sound, and so far down the page that most visitors wouldn't even see it.
Should that be considered a view? NO. Some might argue that an autoplay after a certain percentage should be a legitimate view. I call bullshit. An autoplay video, buried out of sight with the sound off could feasibly play all the way through prior to my ever reaching the end of the page I'm on. That should not then constitute a view. It should not constitute ad impressions on that video as I never even saw the ads or even the video. At GDN, I refuse to use autoplay because I personally hate it and find it to be in poor taste. Yes, even if I visit a video serving site, I do not want video crammed down my pipeline without my conscious agreement. Many of us now use mobile Internet and while 3G might be fast, it's also, in many areas of the world, clocked and paid for by usage. So the embedded autoplay video ot only harms the advertisers who think they are paying for their ads to be actually seen by targeted consumers and the content creators who believe their videos are actually being seen by interested parties but it also harms the very viewers in some cases by costing the money, and for what? So that some unscrupulous individual and make an extra $7 for an extra 1,000 impressions (if the CPM were $7 of course).
What Implications Does This Have For Content Creators?
What does this mean to you as a content creator using these sites? Those views you are counting on should be taken with a grain of salt. What does it mean to you as an advertiser? Well, worst case scenario is that no one is actually seeing your video ads on certain sites and you're paying through the --- for nothing at all. If I were a video advertiser, I might pull any ads I had running against content on those four sites (50%) that count everything, because there is no guarantee that any of the viewing numbers are accurate or legitimate. I might even, being quite a stickler myself, pull ads from all six sites that fell outside of the most stringent category. Really, that's probably what it's going to take to start cleaning up the industry and bring everyone to the table to talk about what really constitutes a valid view.