Go for the ease and efficiency of an ad network, or a one-to-one relationship with a site. What’s the best way to buy online media? In the early days of the World Wide Web, there were only a
handful of properties that were willing to experiment with an ad revenue model and accept advertising. But these cells of advertising didn’t remain alone for long. Very quickly websites sprang up all
over the place that rushed to try the ad revenue model on for size. Within a very short time, there were hundreds of sites that now accepted advertising on their pages. But online advertising was
still in its nascent phase. Chances were that there was only one person who handled an agency’s online advertising efforts…along with a host of other media planning duties that took precedence.
How could a media planner/buyer keep up with the flood of new websites, make contacts, assess them, and then buy them? Enter the Networks. In 1996 DoubleClick offered the bandwidth-challenged media
buyer a one-stop solution to the problem of dealing with the plethora of new sites on the Web. Others like BURST! Media and Petry Networks (which later merged with Katz Millennium and Interactive
Imaginations to form 24/7 Media) also entered the fray to increase efficiencies and give an advertiser reach across a broad variety of properties and thus, put messages in front of a diversity of
audiences by aggregating inventory across a collection of sites that didn’t have the resources or inclination to form their own sales staffs. Like spot TV, spot radio, and national newspaper networks,
the online advertising networks were the answer to prayers the industry hadn’t even started reciting.
But publishers didn’t want to roll over and pass up the opportunity to participate in this new
gold rush that seemed to rest just over the next horizon. Properties like Yahoo quickly gave up their rep firm, Softbank, and formed their own in-house sales staff. Time Inc. formed Pathfinder and
developed their own dedicated sales force. And now, sites all over the Web monetize their pages by selling advertising directly to media buyers—text links, banners, buttons, microsites, interstitials,
etc.
As the industry has evolved, both means of buying media are prominent. Major portals and publishers are self-represented and sell inventory directly to advertisers, just like major networks
and publishers do in the offline world. But the online ad networks also still remain major players in the space, as do rep firms for newspapers, magazines, and local radio and TV stations in the
offline world.
So what is the better way to buy online media? Does the ease of ad networks and the cumulative efficiencies they promise overcome the one-to-one business relationships that can
produce customizable advertising packages proffered by the standalone web site? In this increasingly bandwidth challenged industry, do media planners and buyers get the benefits of service and
attention to detail that should come with single-site representation, or do networks offer enough of what planners and buyers want to negate that? —moderator Jim Meskauskas, Chief Internet Strategist,
Mediasmith, New York
Brian Monahan, President, Inrhythm Marketing: It’s important to distinguish between a rep firm and a network. It just so happens that most of them do both. A rep firm merely
represents a portfolio of sites, while a network to me is where you buy impressions as a commodity—you really don’t care where they run. For both a centralized rep firm with a broad portfolio of
inventory, and someone who also sells a “network buy,” there are three areas I think they add value. There’s an ease of transaction; it’s a one-stop shop, where you have a previous relationship with
the person. It allows you to get things up and running more quickly. Increasingly, the networks have their own ad servers, and what we’re doing with our clients is actually relying on them to do the
trafficking for us. So there’s a level of service that they provide, which eases the transaction. The second point is pricing, particularly when you’re buying impressions as a commodity, you can get
better pricing. And then the third point is the portfolio of inventory. One of the big trends I see out there is this concept of optimization, which is essentially what we’ve been doing for years:
when you have a direct-response objective, you spread your inventory broadly and then you shift your impressions to where you’re getting the most response. Because networks have a vast portfolio of
inventory, they can faciliate that optimization process for you.
Brent Hall, Director of Interactive Marketing Services, iXL: We hear a lot about optimization with regard to the networks. I
believe that in the media-planning process a lot of our so-called optimization happens upfront in the strategic analysis in developing the media strategy. I think some of the advantages that a single
site allows us are things like a clear idea of who we’re reaching. It gives us a better idea of the content that we’re going to be running within. What we’ve seen is that response rates are definitely
higher for single sites, and that’s not only at the click-through rate, but also at the conversion rate. And that largely stems from the fact that it’s a more qualified audience, it’s a tighter fit
with your target, as well as a closer affinity with the content. When it comes to response, mindset is everything, and when you know the content that you’re reaching people within, it’s easier to
drive results. You also get a brand association. Whether our clients actually specify that branding is an objective or not, at the end of the day, they’d much rather prefer running on a site that
they’re familiar with, that they’re comfortable with, than a network where they have no idea where they’re running.
Monahan: You said that strategic analysis happens upfront. I don’t see how this
medium is “plannable.” The research tools that we have talk about sites audience in aggregate, whereas we buy...the media currency that we use does not correspond to a site’s rate base. There’s too
many variables involved between looking at what a site’s audience is—assuming that data’s even good—but looking at that data, and then we buy some section, some percentage of the impressions at some
predetermined date range. It’s almost impossible to kind of hone in on that mindset you talked about.
Hall: Well, I didn’t say that we rely on syndicated research to define our audiences. This
medium is much more intuitive than others, and I still believe that from a psychographic, from a buyer behavior standpoint, there’s a lot that can be derived from doing your homework upfront about an
audience. Back in the day, you’d hear about a lot of agencies doing open runs of, say, the DoubleClick network. Based on that, people would say, “Oh, sports sites work well, or entertain ment sites
work well. So let’s focus all of our impressions that way.”
Monahan: I agree that response rates are higher when you do your homework upfront, but the problem is that most of the time we’re after
cost-per-action. So the response rate can be offset by a premium price, which hurts cherrypicking individual sites. On the concept of brand association, a lot of reps will come in and say, “My site is
a premium site and therefore it deserves a premium price.” I’ve never seen any research that backs that up; I’ve never seen a good definition of what a premium site is. We’re just now starting to get
some insight into whether the latent response—the post-view response—that comes from running on an individual site is higher than the post-click response rate.
Meskauskas: This is a big theme of
mine. That latent—or non-impulse—transaction data that starts to come actually could indeed demonstrate that an individual branded site could in essence yield a more efficient rate of transaction or
conversion rate. If the data does indeed bear that out, what will that do to networks versus sites?
Monahan: This is where I start to switch sides. Right now, I fear that ad impressions on the
Internet are a commodity, which really is going to hurt publishers’ ability to command a premium and build a good user experience. I think one area where individual sites have an advantage is that
they do know their site better and they can customize a package for you so you don’t get just commoditized banner impressions. But—and this is the big but—they need to have a way to deliver their rate
base. Right now, as long as we’re selling impressions, it doesn’t correspond to their group of human beings that actually go to their site. And all the talk about sponsorship is really a creative
discussion. It needs to be a media discussion. We need to talk about the number of unique people and the amount of time and space we get to spend with them.
Meskauskas: Right now, when we discuss
reach on the web, what we’re really talking about is potential cumulative reach. At the Jupiter Conference, they gave out these extraordinary reach numbers for networks, 70% reach or more. In order to
go ahead and achieve that, I’d have to buy every single impression generated by every single site in that network in order to come close to possibly reaching that 70% audience. But if what we’re
talking about is latent or non-impulse transactions, in the offline world anyway, frequency capping is a bad thing for doing that in order to motivate behavior. When a time comes for rate-base-based
ad units—whether that becomes a session length or some other way of cutting it—how is the network going to be able to deal with that? Is there going to be a role for networks if indeed qualitative
rather than quantitative type purchasing takes place?
Hall: The lowest common denominator from an ad unit standpoint with regard to the ad networks is going to harm that in the future. We’re
already seeing the marked decline in click-through rates with regard to banners—banners today are easily overlooked; people understand what they do and where to find them. People understand that they
really aren’t related to the site content anymore. Where the single sites really do have the advantage is in understanding their site, in having the advertising and editorial liberty to go out of the
box, to create customized ad units, to make things that stand out and are more visible to their audience. Implementing new technologies, like rich media, is much easier when you’re dealing with a
single site than when you’re working with a network.
Monahan: As long as we don’t get in trouble with stepping on people’s privacy, I think the networks have the inside track on being able to
deliver a defined group of people a defined number of times, based on cookie technology. But the advantage the publishers have is that they see these people over and over, and they can see how they
use their sites. And if they would invest in some of the pathing and segmentation software that the e-commerce people are doing, they should be able to kind of tie their audience up with a bow and
deliver it to us in a very creative package that a network will never be able to touch.
Hall: In a year or two, we won’t be having this discussion—or at least our definitions of a network and a
single site will be vastly different. The premium single sites out there are in a fight for market share. You’re seeing them acquire competitors and gobble up smaller players in the field in order to
broaden their content and deepen their user base. In essence you’re kind of seeing the channelization of single sites into networks.
To engage in the debate, email debate@mediapost.com.