In 2008, about a billion dollars was spent in online video advertising. In this 2nd part interview with Ooyala founders, Bismarck Lepe, Sean Knapp, and Belsasar Lepe, Bambi Francisco dives into the future of online advertising. In the interview, Bismarck Lepe talks about how Ooyala plans to launch in the first quarter, a subscription capability for its clients.
BF: Let's start with your ad network platform and how you work with advertising networks and advertisers. Can you just explain a little bit how that works?
Bismarck Lepe: Well, right now we are
primarily in the enablement business. We don't have our own ad network.
We have a pretty significant footprint of people who are serving
ads. But I think Sean can probably talk about how we actually plug
into a number of people as part of the advertising ecosystem.
SK: So we've integrated with about half a dozen different ad servers and networks. Everybody from DoubleClick to Lightening Cast to Google AdSense for video as well as Tremor and Yume. We've integrated with both their overlay as well as in-streaming companion ads. And this we offer up for a lot of our larger tier customers who already have a DoubleClick account. For example I can easily plug in their DoubleClick ad tags into monetizing content that way. For those who don't have an ad server account, we actually provide the Ooyala ad server where you can upload your own advertising assets and monetize your content directly through the Ooyala ad server.
BF: OK. So you can go direct [and sell your own ads], which a lot of people do, and play those ads. You can also just use your advertising player, or advertising network, or management system along with the analytics as a separate piece of your service. Correct?
Bismarck Lepe: That's correct. So if we have a repository of ads, one of the offerings that we have is we detect where people are abandoning the video and we make sure not to place ads there. Right and so we try to maximize the amount of time people are going to be spending consuming your content. The more time they spend consuming your content, the more time that you can actually advertise.
BF: What happens if the ad is the reason
they're abandoning the content? Do you take that ad out? To what
extent do the analytics help or aid in determining where to place an
Bismarck Lepe: Well I think advertising usually gets a bad rep. In
general, untargeted advertising is bad. Advertising that actually
complements the content is probably good. If you watch football, Superbowl, the ads actually complement the Superbowl experience.
BF: Like Doritos or beer.
Bismarck Lepe: Exactly. So what we try to do is we try to learn more about the end user. So everyone talks about behavioral targeting. Well I don't necessarily think behavioral targeting works very well in terms of creating a dynamic learning system if you can actually learn from what is causing a certain behavior. So what we do is we actually analyze the video as the video is going through and we extract meta information and then based on behavior we tie it back to what was actually happening on the screen. So if we notice that someone is always fast forwarding through ads whenever there are car ads and whenever they're watching a soap opera then we'll make sure that that mapping of ads to that type of content never happens again.
BF: For that individual person and the way they interact with the ads.
BF Francisco: Do you also take in meta data, tags, keywords and do the publishers provide you with that information? How are you analyzing inside the video? Or are you not analyzing video?
Bismarck Lepe: We do. So we take publisher provided tags for a piece of content and we obviously use that but then as video actually goes through our system, we're also extracting people, certain objects like automobiles, landmarks, logos, and try to learn from that and actually paint a more comprehensive picture of what's happening on the screen.
BF: How effective have the ads been now that you are identifying objects? I was checking out your interactive ads where you
identified a computer and then you popped up a computer - Dell, Apple - ads. How effective has that been to analyzing video. How effective has it been to deliver the right ad?
Bismarck Lepe: Well what you're talking about is actually a different type of experience. We actually expose the APIs for a lot of the technology that is driving the backend of our ad server. People can actually click to purchase experience or the hot-spotting experience that people have been talking about for a very long time. That technology that's being exposed in kind of this hot-spotting experience is actually what's driving our ad server on the backend. So in terms of effectiveness, and using engagement as the bar for effectiveness. We usually see about a 100% increase in the amount of time that's actually spent consuming an ad. So if on average if people consume about 2 minutes of ads during a 30 minute piece of content, we're doubling that. So from a content provider's standpoint, they get more opportunities to actually show advertising to a consumer. From an advertiser's standpoint, they can actually measure the increase and overall engagement with their brand.
BF: Now a billion dollars has been spent on online advertising in 2008, as you've mentioned in the U.S. What's been effective? What are the more popular ads? What are we going to see beyond pre-roll and post roll and overlays? What do you see in 2009 that's going to be embraced more?
Bismarck Lepe: So when we think about advertising, I think people have it
all wrong. It should be about monetization. It's about figuring out
what monetization model works best for content A and person B. If we
know that an individual is actually very willing to purchase a piece
of content, why should we bore them with that? They should have the
ability to pay for that individual stream and pay 5 cents or pay 10
cents. If you think about it. If you build a very simple,
transactional model, where they click once and it's tied to their PayPal account,
and they only pay 10 cents, over a thousand people that's a $100 CPM. So
that's what we're exploring. Obviously, we're continuing to do the pre-roll
and the post-roll and the overlays. And what we've found there is you need
to make sure that the ad is compelling to the end user, the ad
complements the content, but in a lot of cases it's actually better that
you directly give the opportunity for the end user to pay for that
piece of content. If you look at Major League Baseball in 2008, they're
expecting to make close to a half a billion dollars in subscription-based video services. You go to a major baseball live, you actually have to pay for that. And
so i think its never going to be over 50% of the total revenue
opportunity for a content owner, but it is going to be a very real
opportunity for them.
BF: Are you going to provide that functionality for content publishers to be able to charge for their content?
SK: To really build off to what Bismarck is saying, we're in a fundamentally different world online . In the online world, the user is in control. So the traditional in-stream ads for example monetize less. They're conceptually to the user closer to a pop-up ad. It is a very different monetization model than the click to expand. Even take Hulu for example, their option to either watch all your ads up front or spread out throughout the video. Regardless you are giving the user more control at that point. In doing so, you find that you have a much more monetization experience across the board.
BF: So you are going to try to offer that work on building some sort of subscription model in there for them?
Bismarck Lepe: So we're already testing it out with a handful of providers. The results have been interesting. Anytime you go over 50 cents for video on demand content. Content that can be found anywhere else on the Web, the option goes from anywhere like 15%, 20% range down to like .5%.
BF: What are the economics for that? Let's say if Vator wanted to, you know, have a subscription model offering?
Bismarck Lepe: Oh it's great. You just pay us a transaction fee.
Francisco: But what's that transaction fee?
Bismarck Lepe: We're still figuring it out.
BF: What are the economics of the ad? So what goes to you? What goes to the publisher? What goes to the ad network serving of the ad?
Bismarck Lepe: So if it's an ad network that we already support, a content owner who is using features and functionality that's already built in, then we don't charge an additional fee for that. If it's a new ad network, we actually have to kind of integrate into our system, then there will be a professional services fee.
BF:But if you had just purchased just the ad portion of your platform, then you take a fee.
Bismarck Lepe: Then like DoubleClick, we will charge an ad serving fee.
BF: And what is that?
Bismarck Lepe: That's about a $1 CPM. For most of the video ads. Overlays are a little different.
BF: And then, you know it's funny because you talked about the subscription being potentially where the money is in in online video. But is that where the money is in online video? Is it in short form, long form? Where is the money?
Bismarck Lepe: Well I think it's going to be kind of like the old video release or theatrical release of a film where you have windows. Initially a video is shown in a movie theater, then it goes to pay-per-view, then it goes to DVD, and then eventually gets into syndication, then a broader international use of distribution. I think you will probably be able to see the same thing for episodic content. So the season premiere that is going to run in prime time that evening, if you're a subscriber of CBS, you'll be able to watch it five hours beforehand. That evening at the same time maybe you can pay 10 cents per stream to actually watch it without any commercials.Then a week later when they post it, then you actually have the advertising baked in. So it's not an all or nothing model. It's I think a combination of everything.
BF: I think that makes a lot of sense. And when is this subscription, that's really intriguing, when does that come out?
Bismarck Lepe: I will let the guy who owns our product roadmap.
SK: Late Q1 launch.
BF: Ok. I'm sure there's going to be a lot of interest there.
SK: We're seeing quite a bit of interest already.
BF: Well thank you so much, gentlemen for joining me in this segment.