|
|
Media Coverage
"Drilling Down in Behavioral Targeting"
iMedia Connection
May 02, 2005
iMedia Brand Summit panelists continue discussion of benefits and issues of behavioral marketing (Part 4 of 4).
This breakout session took place at February's iMedia Brand Summit in Florida. Sean Finnegan, U.S. director of OMD Digital, moderated.
Read last Wednesday's installment when Dave Morgan, CEO, TACODA Systems, provided some examples of effective behavioral targeting. Thursday, Omar Tawakol, SVP product marketing, Revenue Science Inc., did the same. Friday, the panelists began a discussion on the benefits and issues around behavioral targeting. Here's the conclusion.
Sean Finnegan: Here's one that's near and dear to my heart, as a buyer -- premium. Should this refined and advanced targeting command a premium? Should there be an extra cost of doing business with this? There are obviously two schools of thought, "yes" or "no", so where do you guys take this?
Omar Tawakol: It depends upon what you're talking about in terms of what's your benchmark for premium, right? One view of this is that publishers in the past have been selling an audience. That's what they do, they sell audiences. And run-of-site is a bad word for them. What behavioral targeting allows them to do is to say, well I'm going to give you people who are technology enthusiasts or travelers and I'm going to price that at whatever price it's worth to the market. And is that a premium above run-of-site? Absolutely. Is it more expensive than your best contextual placement? In most of the cases that I've seen in the industry, no, it's actually a discount from your sold-out section. So if you're selling the technology section and it's always sold out, and you sell a technology segment, you might actually discount that. So basically I'm seeing it closer to premium, but at a slight discount. But certainly you're paying more than run-of-site.
Finnegan: Okay.
Dave Morgan: Yeah, I think what's going to happen . I'm of the opinion that rich media has really stunted its growth by its surcharge method of pricing. Rich media is so powerful, but because of the business model it's never been able to be adopted the way it should. We're technology providers and we have to charge publishers for our technology, and since it has limited adoption today, frequently that's reflected in surcharges. But I think to Omar's point, it's just going to be a rate card. There's a rate card for pages. And there's going to be a rate card for audience.
And actually, it's interesting -- a lot of our publishers started, first several years ago, with this sort of middle rate: there's the run-of-site rate and their premium content rate, and a middle rate, which is their new audience targeted rate for that content. We now have a number of publishers because of the back-end performance and the value that are now pricing the audience targeted at the same level, or higher than in-context. And when you can drive sometimes 10 times in branding, or 20 times in conversion, by targeting people out of context -- you find the looker in the automotive section, but you target them not when they're being just inundated with automotive ads, and not when they're involved in automotive research content, but when they're sort of leaning back and relaxed in entertainment content -- you generate dramatically higher return.
But I don't think we can get into just this wholesale rating because what needs to happen is instead of it being used in two or three percent of the campaigns, it needs to be like electricity or television service, always on.
Finnegan: Right. I think the program this morning is running a little late and we're all kind of hungry. But we do have time for one to two questions. Anybody have anything? Sir?
Audience Member: Really two questions. One, you guys started a couple years ago in this kind of interesting timing because that's really when Engage and DoubleClick were shuttering their businesses that were essentially already empowered to engage a product to do this sort of target. So I'm wondering what's changed, what's different now than a few years ago? And the second question is just looking at your numbers . I'm a huge believer in behavioral, but reach is always my problem.
Tawakol: A lot's changed in four years. I know those companies well because I was running one of the ad serving companies that was competing. First, those companies owned the data. And the only companies that ever participated in Engage's profiles network were CMGI portfolio companies. That tells you something.
Most publishers and most clients aren't comfortable with this new massive centralized data owner. So the first thing, it was data ownership. Number two, the market wasn't ready. If you think of all the metrics that were driving the capital markets, it was all about quantity of pages. No one was focused on quality. The New York Times was sort of like the lone one by themselves that was creating a quality audience targeting product; no one was buying it then. Now they are. And now I think that the technology is much better. I mean the technologies we're using, you can't even compare them to the technologies that were being used four years ago.
On the reach issue -- there are two responses I have to the reach issue. I have seen a campaign, a $30,000 campaign that targeted 182 people, generate a return on investment of 10 times. So the world's changing and the idea that you have to go after mass numbers -- and that was a B2B decision makers in health benefits campaign. And that's the best campaign a Fortune 500 company ran last year of any type in ROI. So we have to change the way we think about the numbers we need. We don't have to feel special, that like, wow, I just got a million. I mean, you get the right 200 it can be extraordinary.
But, on the other side, you take a site like About.com -- we've got 35,000,000 plus people that are there. You've got the ability to find those people that are looking for very specific things and then start finding them every time they come back. And it's not a problem to get segments of half a million and a million people that you can generate 50 or a hundred million page views against -- and be looking at pure quality.
Morgan: Yeah, the reach comment you made is an important one. I've seen some of our auto publishers extend their target-able inventory by 50 percent because they're able to do behavioral targeting. And reach is on their mind too, because the last thing they want to do is convince you to buy a new type of ad package and then sell less volume. If they sell 10 times less they're not going to charge necessarily 10 times more. So reach is on their mind too.
The only thing I would say is that a technique I'm starting to see gain traction is aggregation of many segments. So for instance, again, an auto example, they were selling a campaign to a particular auto advertiser, and they were able to create literally 30 different segments that went into it. And so they're really approaching it from a bottoms-up perspective, aggregating highly valuable niches that, when you add them up, give you good volume, as opposed to starting out from a top-down approach, which is this huge segment of the site that's un-targeted.
Finnegan: Great. All right. One more.
Audience Member: This morning Poindexter talked about behavioral, and they mentioned both of your companies, specifically. And my question is: how do you guys work with them because they talk about how they're targeting the marketers and the agencies, where you guys are targeting publishers, that kind of thing. How do you guys work together?
Tawakol: Well, they drive the buy side and we drive the sell side. So imagine that a financial services company wants to do more targeted acquisition programs, they use Poindexter to manage their data and their databases to better optimize their creative. And TACODA and Revenue Science provide the tools to the publisher so that when Poindexter and the financial services company and their agency say, I want just these people, they can be delivered up.
Over time I think we'll have a lot of participation among ourselves on coming up with some of those standards that Sean talked about.
Morgan: Yeah, exactly, the buy/sell side distinction. We're working with you guys. And you guys now are able to really understand your audience. But the key is that you now own understanding of your audience and you sell it. What Poindexter can do -- and of course this rubs a lot of publishers the wrong way -- is in some cases they can strip mine or cherry pick. And they're sensitive to that now. They're a smart crew. It is a fundamental problem because once they go in and do a campaign on your site, they're cookie-ing them, they own that database, and they want to leverage that information as opposed to you.
Audience Member: It seems like there would be a little bit of a conflict with the two systems and the cherry picking of the impressions and the people, compared to the methodology that you all are tying to implement. I'm not saying anything bad about Poindexter. I just don't see the two working together.
Tawakol: One of the ways to look at what we do, very frankly, is we're gatekeeper systems. Publishers realize that people have been going onto their sites and buying inventory for 50 and 75 cents and converting it into $5.00 stuff after clicks and acquisitions and reselling. And they want to control that margin. So they realize they need to optimize their inventory for the hand-off. We're basically arms' merchants to different sides of the game.
Morgan: The last thing I'd say there is that I do think that in the future there is going to be some merging where advertisers may be able to come in and say, here's a segment of people I actually see from my site, so not just an RFP, an actual set of cookies.
But we're not there yet. You now have two separate sides and they're not talking to each other.
Finnegan: Okay. We could probably spend a lot more time because nobody has left yet, so we see you're half-way interested. I do appreciate your time and interest, and let's give a hand to Dave and Omar.
[Applause]
<< Return to Previous Page |
Privacy
Site Map
Home
![[x+1]](../images/logo_xplus_main.gif){kind=logo}
