From the Department of Duh:
A recent study by IBM, titled “The End of Advertising as We Know It,” found that 40 percent of the 2,400 consumers and 80 advertising executives it surveyed found ads during an online video segment more annoying than any other format.
Particularly egregious are pre-roll ads, those 10-to-30 second video spots shown before the actual video, as any Web video entrepreneur can tell you. Mid-rolls (spots in the middle of a video like a regular TV commercial) and post-roll are other variations. But there are lots of alternatives—traditional banner ads around the video in the player itself, clickable ad “bugs” that crawl across the screen, pop-up ad-overlays, even video hyperlinks that make people and objects inside the videos linkable.
A lot of this is hit or miss. Take the idea of video hyperlinks, where you can draw a little box around an object in a video or highlight them in some other way to signal that you can click on it for more information or to see an ad. Below is an example of how that looks like from Asterpix (other startups, as well as Microsoft, are working on similar technology):
While the idea is cool, I’m not sure it works. The box is too distracting and ruins your enjoyment of the actual video. Links in a separate sidebar would be better.Advertisers like to stick with pre-rolls because they are easy to make and it is a format they understand. But as the more adventurous among them experiment with different formats, though, they will gravitate towards the ones that actually produce the best results. Tell us what type of video ads you think have the best chance to succeed in the poll below:
- Contextual text ads on the side
- Something else
- Clickable ad “bugs” that crawl across the bottom of the screen
- Display ads on the side
- Video Hyperlinks
- Banner overlays that pop up over the video
- Video Pre-Rolls
- Video Post-Rolls
- Video Mid-Rolls
Total Votes: 1032
Started: November 9, 2007
I am a fan of overlays that give you the option of clicking through to see the advertising. I think YouTube was close to having it right with that experiment that they ran a few months back.
The problem with videos with banners around them is that they end up being extremely large and extremely ugly. No one wants to have to find a place to embed some non-standard video player on their blog/website etc… I think this is one of the reasons that YouTube’s video ad unit hasn’t really caught fire.
Pre-rolls are annoying, post-rolls never get watched, the only unit that seems to make any sense at all is one that doesn’t increase the player side and allows the viewer to opt-in.
ABC.com, Might have the best option to distribute the commercials through the show. I don’t like it, but I do see the commercials. My girlfriend and I make a game out the “click to continue” sometimes I forget “because I’m watching the ad” I’m not sure if its the best, but it works. Good job to the team that made the ABC online video system.
All the ad unit formats would certainly distract viewing experience, that’s why i think this has to be done in a whole different way than thinking about where to put the ads.
How we can seamlessly make ads being part of the viewing experience, rather than just contextual or related links.
Also, maybe you can add this to the poll: Ads displayed when user pauses the video and it disappears after unpause.
That’s pretty creative Yang. It might be hard to get good estimates on impressions though.
I agree with Steve about overlay ads. I think they work best for short and long form videos. From the advertisers point-of-view, they might not be the best because viewers will simply ignore them. Some sort of context sensitive overlay ad would be really cool!
For long video, like ABC.com, I’m okay with the short mid-roll ads (NEVER pre-roll). On ABC.com, the ads are 30 secs or something like that. When watching a 40 min show, you might get two or three 30-sec breaks. 1.5 mins of ads for 40 mins of quality tv seems reasonable.
Also, I think the quick “this video is sponsored by” blurb is okay for a pre-roll, but keep these short (under 5 secs).
I work at an instructional video (fitness/wellness) site http://www.beYOU.tv and we are experimenting with different types of ads. Currently we are running mid-rolls, but I’m sure we’ll move to something less intrusive in the future. Mid-roll ads work well with instruction content. We break our videos into segments, each segment is a self-contained clip, for example the proper push up form, would be a segment. Once the viewer has watched the segment, they (in theory) have gained the knowledge that they were looking for and when the ad plays, it is less disruptive.
If you are working with instructional video, how do you run ads? I’d love to hear your ideas!
Greg
Anything is a small price to pay for all the free entertainment and news on demand.
Users should really be more considerate of those Advertisers.
It takes so much to get them to sponsor a site – how does it feel when they read these criticisms.
Do we NOT realize that without them, most of this expensive bandwidth would not be free – or it would be so compressed and pixelated. (remember YouTube two years ago)
That was the sickest surfing video that I have seen in a long time.
As far as video ads, this is the $100 billion dollar question.
I think that 3 to 5-second pre-rolls are the way to go followed by another post-roll video, or interactive flash, or graphical ad with a link, all of which would be by the same company.
Ie. 5 second pre-roll of a BMW M3, followed by a display ad by your local BMW dealership asking you to call them or visit their website.
Back to the replay of the massive wave…
It really depends on the content and the context. Most online video is still ’short form’ at the moment. In a year or twom when everyone has a media centre, more content will be long form and this require a different advertising approach again.
The site I work for (http://www.wedigtv.com) serves over 100m 30second mid-stream ads a month and these seem to go down pretty well with the user, but that is largely (I think) because they are in keeping with the content.
IMHO it’s going to be branded content, product placement and the like that comes more to the fore when the PC replaces broadcast TV in the sitting rooms of the world.
Hey steve.
That is indeed a problem with ads during pauses. Since you cannot control where viewers pause…
Maybe advertisers can pay based on the number of times that their ads were displayed when users pause.
This will be a better model for video sharing sites that focuses on “shows”. I know that i pause quite often in long videos =]
That would certainly be an interesting experiment Yang. I would love to see what the pause numbers look like for really big video producers and whether it would be enough to make it worth doing.
Mid-video commercials are what work best on my site for our advertisers. Pre-roll sucks because it’s like asking someone to pay for something without giving them anything. I don’t know about everyone else but if there is a pre-roll I usually click that little red “X” in the corner. I think that advertiser/publisher determined mid-video is probably the hardest to implement but the best for viewers.
Personally I prefer midroll. One datapoint though, if the site is using Flash Media Server to serve the video, the player can indeed signal to the server that the content has been paused. I haven’t seen anything that uses this functionality, but I doubt throwing an ad script in that equation would be difficult.
I’m in Interactive advertising, and this is indeed the million dollar question. I have talked to ad agency people (interactive and television) and everyone has a differing opinion right now, and often they are too busy fighting in house about who gets to monetize on-line videos (the traditional television side of the house or interactive).
Pre-roll ads are a joke- and everyone hates them. No news there. IMO, mid-roll is the only way to go. The boxes here are interesting, but ultimately annoying as hell. I got sick of them after about 3 seconds, and they distracted from the quality. Mid-roll would be pointless in a video this short– maybe at the end of the video, it stops and darkens with the box highlighting just the surfer. When the user moves over to see why he is highlighted, they see the link, and hopefully click through.
The way this was presented is just going to annoy users to no end. Mid-roll for longer videos- for short ones the jury is still out……..
ooyala.com is doing stuff with this as well…seems cool. this shows some of what they’re planning to do i guess. http://www.busi...com/article/248 is a good interview about it and they have been recruiting a lot at stanford.
found a link but it doesn’t talk about advertising yet… FAq page says it’s coming http://www.ooyala.com/backlot
One company that is missing from this article is VideoClix. They provide Video Overlays that display product information and links to eCommerce backends or informational sites when any object is clicked on in a Video. Also, as Erik mentioned in the article, VideoClix also allows object information appear on the side of the video so it does not disrupt the Video image. At the same time all of the clicks that happen are tracked for the user so they can visit what they are interested in after the Video is complete. This allows for a targeted, “lean forward” advertising experience for users as they only get information on what they request.
You can check them out at Http://www.VideoClix.com
Anything that doesn’t make you wait is better. Making someone wait is the only 100% assured way to frustrate.
Completely agree with Greg.
I wrote an article on my blog about the video pause ads I talked about earlier if anyone want to discuss.
http://www.3vitv.com
i like maybe something like bottom of screen is an ad
Erick
Thanks for the feedback on asterpix. You can select the indicator to suit your tastes using the style button, ranging from (a) the “bold rectangle” that is the default for first time viewers who are otherwise unaware of the links in the video, (b) muted rectangle, (c) intermittent beacon, (d) or no indicator. In all cases, the logo on the top right of the player turns green when a hot spot is present as a visual cue and the time bar indicates the span of the hyperlink. So you don’t need to compromise on visual enjoyment to enjoy nonlinear navigation of the video. I might add the links in the video can be to other videos such as in http://www.aste...ole?avi=7695561
Best,
-nat
There are a couple of companies that are bringing Overlay technology to the next level. The best of these I have seen is a start-Up called Panache http://www.panachetech.com They are powering some of the bigger players according to their VP of Biz Dev. Their tech allows the insertion of any interactive experience on top of a video. There are some really great demo’s of this on their site.
The other company is out of the Netherlands and has similar technology – I forget their name but saw them at AdTech. Both of these companies are not trying to define the format but instead offer a way for both the advertiser and the site to control how the message is delivered in very creative ways.
The future of Video ads is not in the format but in the delivery of the message with each site/channel allowed to determine what works best for their viewers and advertisers. With integration into the video and real time dynamic ads as the bingo the concept of Rolls will fade into the past as interactivity is developed inside the video itself and the message becomes the medium.
I think a good way to enable short form video to pay for itself is to make it commercial to begin with. I’ve spent a significant amount of time watching ads and real tv on youtube, and a postroll on the tv that says when and what channel I can watch would not have been intrusive. Along the lines of commercials, perhaps video sites could show commercials in the related videos crawl. The trick is that they’d have to be as entertaining as the best superbowl ads, all year round.
Remember back in the day when TV was aired live all the time because there was no tape and announcers would cut to a sponsored plug? yea, that seems to work the best.
Wow, all of these super complicated ideas…
Why can’t we just get back to the basics? Why not just take the AdWords concept to another level? I mean, it seems to be working OK for text content.
Stick a 30 pixel high “bar” at the bottom of the video (longer is better) and display simple, contextual text ads. Is the user watching a Star Trek clip? Tell them that there are Star Trek DVDs on sale somewhere for mighty cheap. Did the video just change to a scene with a popular actor? Change the ad to something relevant — if he’s got a new movie that’s about to hit theaters, link me to the site. That’s advertising, but it’s *useful* advertising.
I don’t understand people wanting to foist 1990 methodologies into 2007’s digital age — it’s like freaking *time-travel*, as if they completely skipped the information revolution. Who didn’t notice the arrival of contextually useful information in advertising on the Internet?
It’s like, I’m watching the Sci Fi channel and a commercial for Tampons comes on. Yes, Tampons. During the commercial break of Stargate: Atlantis. Really? REALLY? There’s so many other things they could’ve showed me instead, but they picked Tampons. That is the very manifestation of advertisers just not paying attention to what they’re doing at the moment.
I in no way shape or form am going to click on irrelevant *information* in the same way that watching a Sci Fi flick is going to make me want to bounce to the store for a Maxx Paxx of Period Stopping Power, because its *irrelevant* and *stupid.*
Advertising should at least be tangentially related to whatever the user’s doing or watching.
Hell, TechCrunch is a good example of that — look at the sidebar. Ton of ads, but they’re all relevant to the readership.
Wow, rant off.
I might be playing devils advocate here, but pre-rolls and post-rolls are still the most effective forms of advertising, even if they diminish the user experience. Lay-overs come close, but you can still ignore them where as you cant ignore a pre-roll for example. Sometimes advertising comes down to being annoying so you actually get a viewers attention.
AW,
The difficulty of gathering contextually relevant data from a Video is greater than you think. It has to be garnered in some sort of automated manner to produce a relevant ad. This is not easy when you are talking about Short Form User Generated Content and requires human intervention (tagging) to assure editorial integrity. Long form professional content is a bit easier but is not as much of a issue since people are used to commercials during their 30 min program.
There are many start-ups trying to tackle this problem: DigitalSmiths, ScanScout, AdaptTV to name a few. But its neither foolproof or easy and with failure rates about 20% branded advertisers will not risk their ad appearing over content that could be damaging.
The solution is far from here so in the meantime people are looking for a viable solution for the obvious loss of ad dollars in the video.
@ 25:
I actually ignore pre-rolls all the time. This isn’t the TV watching experience we’re talking about here– I just hit mute and tab away (Firefox), then come back about a minute later when I remember (if I remember), then move the progress bar back to the beginning of the movie. I suspect a lot of people do that.
Well, I used to hit mute. Since I’m on Vista now I just kill the audio from FireFox until I’m ready. Takes about half a second since i keep the mixer open. Keeps my media player humming along happily without some guy blaring on about something I don’t care about.
@ 26:
Hey, I’m not tryin’ to super-hate or anything. I’m not even saying it has to be 100% automated, just (tangentially) relevant.
I mean, hell, Google could set up some Mechanical Turk tasks: watch a video broken up into random lengths, and have the Workers apply tags to whatever content is present in those 5, 7 seconds. Might cost them a few cents that they’d make back in less than a day by people viewing the video. Every advertising-supported video that hits YouTube would have enough data within hours for them to target with relative accuracy.
I’m not in the advertising business or anything — to me, it looks like a problem that could solved by existing technologies, rather than having to rely on complicated overlays (like, wtf is up with that player embedded in the post?!), or annoying pre-roll / post-rolls that doesn’t even guarantee a view for the annoyance factor.
Sorry if I sound mildly deranged: I just don’t think it’ll work in the long run for digital content when alternatives are a click (literally) away for end users.
If you’ve ever found yourself going to YouTube because it takes too long to find a video (or for the video to start) on another site, you’ll know what I mean.
Check-out brazilian start-up Boo-box.com… they’ve just released some contextual ad tools for videos
This isn’t a million dollar question. It’s a $70 Billion question. Will web video advertising ultimately provide at least as much value to traditional brand advertisers as they get from broadcast TV? We know the TV advertising model is broken. We don’t know if web video can replace it. So far the results are discouraging…to advertisers.
I think the hyperlinks are a great idea. Pre-roll advertising I hate. I can live with post-roll, but needs to be done in a tactful way. http://www.crenk.com
I had not come across Asterpix.com before and I’m wowed by its potential. That is just the tightest targeting I have ever seen. Imagine if the link over that jet ski there had been a link to the distributor’s product page – what better endorsment could there be than that it’s used to haul suicidal lunatics onto waves like that? The hotspot indicator is a little intrusive, but it could easily be reduced.
The other models you have looked at are all interruption advertising of one sort or another, and inevitably users will develop strategies to block or ingnore them. I would expect to see a much, much higher clickthrough rate for Asterpix’ model.
Interesting
http://www.meetingflex.com
One of the things I spoke about at the recent Video On the Net show (http://www.vide...#hwsf1177584754) in Boston was the need for innovation with online video. The whole “Video Advertising” industry is based on one thing, “What is the best/least way to upset a viewer’s experience”.
The host on the panel, Anton Denissov, Analyst, Yankee Group, mentioned that there are 4 ways to monetize video now.
1. Advertising (which I call the necessary evil)
2. Sell-Through – This is Qoof’s focus
3. Subscriptions
4. On Demand/Rentals
I believe the industry is too focused on #1 and there is not enough innovation on the others. In addition I believe that in the next 12 – 24 months we will see the list of 4 grow to a list of 12. The innovators that come up with new ways to monetize video will have the greatest chance of success. It is time to start thinking outside the “advertising TV Box”.
Come on guys! Let’s get innovative!
Pre-rolls do not suck, that’s my favorite part.
http://fakestev...er.blogspot.com
Tsk, tsk, tsk. All this brain-storming just for reinventing the wheel. TV has had several decades to perfect video ads, so video ads will be the same as TV ads. The only wrench is clickability, but it does not make a huge difference because once you start staring at video, you tend to forget about your mousy.
As #5 (Greg) said, ads will become more long-form, and consequently they will get closer and closer to what the terrestial networks are dishing out today.
@35
I think you are right but still does not make it the most efficient way on the web.
contextual contextual contextual.
when a company nails search within and displays ads on the fly matching the contents, it’s game over. that is the future and theres a lot of companies working on it.
Prerolls do suck. I have no idea how many people do this but
1) if there is a “skip this ad” button I click it immediately
2) if there is no “skip this ad” button I just decide not to watch the video I was originally interested it. I close the page.
The solution that we see is a value exchange ad model: agree to an upfront, honest exchange of your attention with a full-screen, two-way interactive ad to earn video content.
With this type of solution, you can grant access to 30 minutes or 1 hour worth of all types of video streams. When that time expires you watch another ad, not when you switch to another video. By giving control to viewer to start, interact with and complete the ad at there own pace, you ensure the advertiser (who pays to keep content free) that their ad is engaged by the viewer.
Other ad formats compete for attention over and around video or simply appear without consent to the point of distraction and frustration. We believe making an explicit exchange:
- Respects viewers who we hope are savvy enough to understand the value exchange to keep content free.
- Gives advertisers guaranteed engagement especially when they build interactive ads and don’t simply put up a :30.
- Gives content owners a flexible format that separates ads from video in a honest, opt-in format.
Paul Grusche Ultramercial
Had raised this issue in a forum Alsnmaiip on how the video ad and how much. I saw many of the way he failed as advertising what they said. As it makes visiting the injured Palmml download messages from slow browsing. From my point of view I do not see any benefit