NTV asks what the big ad format will be in online video. No real surprises in what the panel answers. Can’t blame them.
I think the dominant ad format will not be pre-roll or overlay, both are somewhat intrusive:
- The pre-roll is like the pop-up. How much longevity did that have?
- Any decent quality video will have some kind of caption in the lower 1/3 of the screen, and the overlay blocks it.
Much the same way Google (technically GoTo.com first) took simple little text ads (classifieds basically) and deployed them online, I think what will succeed in online video will be a throwback to something from offline media, this time TV.
What is that?
Just a guess, but what about the PiP. That’s right, the Picture in Picture. I think it will work for a few reasons:
1 - a PiP ad format can technically stay on throughout the whole video, which I think is the only way video will surpass search advertising’s size. I am not saying it has to remain throughout the whole period, but that would be far more effective.
2 - a PiP can be an image, short text or even, video. If video is the killer app, should the killer ad format in the killer content format not include itself? Expecting a video ad format to thrive with no video in it is akin to building a website out of a cardboard box and calling that interactive.
3 - Technically, you should be able to somehow make it expand to see the whole ad, so imagine if
- scrolling over with a mouse will make the PiP expand as an overlay with the full text;
- a simple click will make the PiP load a video (I know, this might be irritating) and
- a double click takes you to the client site.
It’s not obvious but it could work, especially as viewers learn how the PiP works.
4 - Video content itself has expanded, at WatchMojo.com, for example, we publish in 480×270 pixel sizes, so a PiP of 120×90 would not really piss off too many users, I think… look at the image below. I’ve noticed on our page on YouTube, for example, that an overlay does cover up the caption that bears our text, so I do not think that is optimal, either.
5 - I don’t think it’s very killer (in a good way) for the overlay to have to go somewhere, such as the lower 1/3 of the screen, whereas technically the PiP can go to the top, bottom, left or right. With our content, for example, I’d probably put it on top, to the right.
See the black box, that’s what a 120×90 PiP would look like in our 480×270 publish format.
That black box does not really take much away from the content, does it? Then again, maybe I’ve had too much turkey this week, cause a Black Box is obtrusive… but a logo, text or image would be easier on the eyes.
I’m not saying no one has done this, I’m just saying no one has really fully executed this… and this is where I think the most upside is.
What do you think? More importantly, if you wanna watch the video in question, click here, it’s pretty funny.
Subscribe:
December 28th, 2007 at 9:47 am
Hello Ashkan,
I’ve been an avid reader of your blog for a while now and I’ve agreed with all your positions. Well all, except this one.
I think PiP is very bad idea. I think it is highly intrusive and it will take a lot from the video. The example above shows a black PiP that given the use of black in the underlying video, could very well be part of the underlying video. In practice, the PiPs will contrast with the underlying video and will make users wonder what they are missing, and therein lies the problem.
Another way to look at it is to ask oneself this question - would I like to watch regular TV with PiP enabled?
Furthermore, for courtesy’s sake, I assume the video players will allow viewers to close the PiPs and if so, I think most people will close the PiPs immediately. Alternatively, the PiP could be programmed to fade out after 10-20 seconds of inactivity but I don’t think that approach solves the problem.
In my honest opinion, I think videos should have no overlays, no pre-rolls and no PiPs. Instead, I think the websites that show the videos should detect the context of the video and show one or two relevant text ads (or video ad thumbnails) as a caption underneath the video.
For example, a video that shows the Spice Girls could include ads to buy Spice girl music tracks online, or Spice Girl memorabilia, tickets to the Spice Girl reunion concert, or could show thumbnails to other videos from their record company.
For non music video ads, the same concept applies, identify a dominant commercial product or concept within the video and show complementary ads OUTSIDE the video but CLOSE enough to allow the user draw an explicit link between the video and the external ad.
We should focus on getting a relevant message across not necessarily matching a video creative with a video or trying to be more intrusive than necessary.
That’s what I think will work anyway.
As a disclaimer, I’m part of the team that manages the Aductions Advertising Platform. We currently don’t support video ads explicity but we will be doing so very soon.
December 28th, 2007 at 9:55 am
Thanks Niyi. I hear you loud and clear and you raise excellent points. Maybe the PiP has no future, but, when you say:
===
In my honest opinion, I think videos should have no overlays, no pre-rolls and no PiPs. Instead, I think the websites that show the videos should detect the context of the video and show one or two relevant text ads (or video ad thumbnails) as a caption underneath the video.
For example, a video that shows the Spice Girls could include ads to buy Spice girl music tracks online, or Spice Girl memorabilia, tickets to the Spice Girl reunion concert, or could show thumbnails to other videos from their record company.
===
I do not think any CPC-based, performance based type of ads will really work in video, the reason being that when we watch video, we press Play, then lean back - hands off the mouse - and watch the video. As such our propensity to click through any ad is close to nil.
Video ad formats will have to be display and branding oriented to become as large as we say they will.
CPC ads worked in search because you searched for a keyword, and your next action was to click on something, be it on a site (organic) or a paid ad… but with video, your hand is usually not on the mouse, and as such, CPC-based metrics will fail.
December 30th, 2007 at 6:32 am
Hello Ashkan,
Good point. I agree with you that CPC campaigns work best when you can show ads when users are actively trying to find information.
However, the “caption ads” that I tried to describe above aren’t really meant to be conventional PPC text ads. They could very well be brand ads. The most important point that I wanted to stress was that the ads shouldn’t interrupt or be overlaid on top of the video.
Let me expatiate on how these ads could work:
Imagine a scenario where all video distribution channels were also transactional platforms. By that, I mean, viewers would be able to initiate a purchase of a product or a subscription to a service through those platforms.
For example, if I was watching a clip from The Family Guy on YouTube, it will be infinitely more efficient if I could also order a DVD or the full episode by clicking on a button (in the caption ad). One click, $1 and the full episode is mine. Very efficient.
I believe that to successfully monetize video, content owners and distribution networks will increasingly have to think of clever ways to CROSS-SELL other services DIRECTLY to the view. It could be other products, free branded t-shirts, essentially anything that complements the video.
It’s the classic Razor and Blades model, if you will.
December 30th, 2007 at 6:32 am
Hello Ashkan,
Good point. I agree with you that CPC campaigns work best when you can show ads when users are actively trying to find information.
However, the “caption ads” that I tried to describe above aren’t really meant to be conventional PPC text ads. They could very well be brand ads. The most important point that I wanted to stress was that the ads shouldn’t interrupt or be overlaid on top of the video.
Let me expatiate on how these ads could work:
Imagine a scenario where all video distribution channels were also transactional platforms. By that, I mean, viewers would be able to initiate a purchase of a product or a subscription to a service through those platforms.
For example, if I was watching a clip from The Family Guy on YouTube, it will be infinitely more efficient if I could also order a DVD or the full episode by clicking on a button (in the caption ad). One click, $1 and the full episode is mine. Very efficient.
I believe that to successfully monetize video, content owners and distribution networks will increasingly have to think of clever ways to CROSS-SELL other services DIRECTLY to the viewer. It could be other products, free branded t-shirts, or essentially anything that complements the video.
It’s the classic Razor and Blades model, if you will.
January 1st, 2008 at 9:58 pm
[…] Internet Video is the Granddaddy of internet trends. Ashkan looks into what video ad format will work . […]
August 18th, 2008 at 2:05 pm
[…] I don’t see anything else becoming a category killer like AdSense in search (what about the PiP?). But what if video became a cost - or marketing expense - in order to have more content to […]