Disaggregated Media (continued) - The Rise of the Ad Networks

Yesterday I posted on the subject of the disaggregation of the vertically integrated media model. In my haste to to finish the post and head off to dinner with friends, I left out probably the most important part of the system - the revenue producing function.

Just like content creation is happening separately from editorial which in turn is happening separately from distribution of the content, I believe the ad sales function will be largely divorced from the production and distribution of media.

That's not an obvious outcome and certainly all major internet media businesses (from NewYorkTimes.com all the way to MySpace.com) have direct sales organizations to monetize their pages. So why do I think the monetization function will largely be divorced from the creation and distribution of content online?

Well first, we can see it happening already. Search is a classic example of this. The first step in the content consumption process (finding the content you want) is being monetized by Google and Yahoo! and others, not by the companies that a producing the very content you want to find.

I spent some time this morning looking for the 2005 online ad market data broken out by the key categories but I couldn't find it. If and when I do find it, I'll update this post with the numbers. But search probably represents about 35-40% of all online ad revenues right now and none of that is sold by the media properties.

The next fastest growth category in the online ad market is third party ad networks. They've been around for a while, really since the beginning of the Internet when Doubleclick and others offered to sell advertising for their customers. Those early entrants are largely gone from the scene now, but they've been replaced by ever more complex and effective ad network approaches.

In the bleak years after the first bubble burst, the "remnant" networks rose to prominence. They bought or simply took on the role of selling all the ad inventory an online property couldn't sell themselves. And in those years, that was a large portion of the total inventory a site had to sell.

Companies like advertising.com, fastclick, valueclick, burst, and others aggregated that inventory, did some segmentation, and performance monitoring, and sold it off in bulk at CPMs between $0.25 and $0.50. And they built very large businesses doing this.

Last year, these networks represented at least 20% and possibly as much as 30% of the online advertising market. And I suspect they continue to grow more quickly than the portion of the online ad market that is sold directly by a media property's own sales force.

But we are now witnessing the rise of "premium" third party ad networks that do more than just get rid of remnant inventory. They often are able to generate effective CPMs approaching those that a direct salesforce can generate. The contextual networks like Google's AdSense and Yahoo! YPN are the best know of these. And for certain kinds of content, like a gadget blog like Engadget or Gizmodo, or a highly targeted healthcare site, they are likely to result in CPMs that approach what a direct salesforce can get.

Now were are seeing the next generation of ad networks like behaviorally targeted networks (we have an investment in the leading one - TACODA), video ad networks like Tremor and Lightningcast, feed powered ad networks (we have an investment in FeedBurner), lead generation exchanges like Root and LeadPoint, and many other emerging premium third party ad networks.

These premium networks use scale, reach, and data to deliver better ad performance than any single web property, with the possible exception of the portals, can offer on their own.

Today, most ad servers have a "waterfall' algorithm that puts ads that are sold internally at the top of the ad server rotation, then they move down the value chain, measued by effective CPM, in order to move all of their inventory at the best available price.

I predict that within the next couple years, we are going to see these algorithms start putting some  third party ad networks above the internally sold ads. And in five years, we will see some major online media companies starting to question the wisdom of having an internal sales force.

This doesn't mean that ad sales people will be out of jobs, it simply means that they will end up working for companies that specialize in the revenue producing function, not the content creation function (of course if you read the previous post, those companies are going to look a lot different too).

Just like the computer industry went from a vertically integrated model 30 years ago to a horizontal model today, we are seeing technology force that change upon industry after industry. The media business will be no different. Of that I am sure.

Comments

[Adam Elend]

I agree that disagreggation is happening and will continue to happen, though your cable example of cnn can be broken down one step further: Independent content production companies produce shows for networks like, say, Discovery Health. Discovery then pays the Time Warner for space on the network.

A point I think you missed in your posts, though - The commodity that disappears as media moves online and on demand is time and space. The real estate that Time Warner sells to Discovery is no longer valuable because a) they no longer have exclusive or near exclusive access to that geographical region, b) search means that much, much more content that can fit in one space (or within one media network), and c) the relationship between airtime and ad dollars that used to be the main source of revenue for the middleman (cnn in your example or discovery in mine) becomes much less important.

What replaces these elements on the web and on demand is traffic. Traffic drives all revenue, and without restrictions in time and space, traffic becomes much more difficult to come by.

That's why my prediction is that content will shift from a "broadband channel" model, where you go to a specific spot on the web to see the content you want, to more contextual syndicated distribution - just like the advertising models you describe in part two of this post.

Google is already now incorporating video ads into it's adsense program. Forget the ad part, to me that just seems like a perfect vehicle for distributing high quality content in places where the traffic already exists.

Whether the monetization of this model would be the reverse of adsense (websites paying for high quality entertainment content in the same kind of a bidding system) or would be more of a shared model (content producers pay for the space on the sites just like adsense, and then sell embedded advertising) I don't know.

But either way, the end result of this model would be media networks that work like ad networks - where you would sign up for the X network and pay for the high quality X content you know and trust to be syndicated on your site contextually.

Great jumping off point for a discussion - I also take issue with your assertions about user generated content... there is a reason that shows like 24 cost 2 million an episode to produce, and a reason we will continue to consume that kind of content at a much more meaningful level than the firefox ad, no matter how clever and professionally produced it is. But this is already too long, and I'll save that for another time.

[ben barren]

The growth of networks, ad algorithms, self service technology etc i totally agree with. However many of the new ad technology players are also looking to enterprise license their technology (eg ads-click.com) so that media and online publishers with large salesforces and advertisers, can keep "their" revenue, and then just pay for licensing of the technology.

if im a local newspaper group with 30-150 newspapers for example and a large ad sales force with rships with local advertisers, it makes more sense to get the best of new ad technology (that can deliver local ads, optimise, self service, billing etc) but maintain the ad rships that I have had for decades (and more)

as u said, its a 5 year trend this one...

[Colin Crawford]

You can find the FY 2005 breakdown of Online Revenue from the IAB/PriceWaterhouse Coopers report at http://www.iab.net/news/pr_2006_04_20.asp which shows search at 41% of all revenues. Google is currently estimated to be taking 25% of all US Online revenues.

One revenue stream you've not focused on is lead generation - while currently is seen as mainly a B2B online revenue stream but I believe this will spread B2C items. Many B2B publishers will make more from this revenue stream than straight CPM based advertising becuase of some of the reasons you've identified. Already Google is testing Click Per Action which is obviously a form of lead generation however lead generation involves more than just sending in an inquiry - there is also a lot of lead nurturing and profiling required if the maximum value for the lead is to be realized.

[Michael Cooper]

Pushing against the abondonment of your own sales force is the fact that these networks optimize their revenues, not yours. They do not tell you how to appeal to advertsiers .. they use your content when it makes sense for the advertisers they have in their stable. There is yet "another way".

[Bennett Zucker]

Another great and insightful post. I would add that most great branded content sites have both high-value brand-building inventory and lower-value, non-premium inventory, ensuring bright futures for both networks and site sales teams.

Non-premium inventory can indeed be commoditized without harming the brand or the sales team's efforts. With so many 3rd party networks competing for non-premium inventory, publishers have the option of putting them all on a level playing field to compete against each other in fair and open exchanges that are now open for business.

Networks that add great value by offering superior optimization engines, behavioral or contextual targeting, etc. will rise and the rest will likely fail. Meanwhile, great site sales teams will be free to become even more focused and better at selling high-value ad opportunities.

To Ben Barren's point, the ability to run a fair and open exchange that puts all networks on a level playing field is very much available to large sites and proprietary networks as well through the same mechanisms. Once again, this ensures that both networks and site sellers should prosper.

[Tech Guru Adam]

I agree with Bennett Zucker's comment. Internet Advertising should be sold through an exchange that allow fair play; level playing field. It is obvious each ad space carries a different ad value to different advertiser. So why not let them determine the real ad value of each ad space. This is why a platform or a main hub is necessary to allow the media buyers do just that. The only ad exchange that comes close to what I am describing is Hyperbidder, a relatively younger ad network that is getting a lot of buzz lately, and deservingly so. Their technology is unique in a way that they allow advertisers decide how much they should pay for any ad space on any web site listed. They let the demand control the frequency of an ad being served based on the advertisers' rankings which are determined based on their bid amounts. I believe we will see big changes in the ad delivery and sales in the next five years, and the leader will be the one that can win the hearts of both sides; Advertisers as well as Publishers. So far, I only saw one company that is well positioned to do that. I believe in the future, ads will be sold more transparently on a level playing field, and that the need for internal sales team will start to diminish.

[nick gogerty]

hey how about user generated content streams from google video like this...
http://nickgogerty.typepad.com/thoughts_for_now/2006/06/google_video_rs.html

[Jim Larrison]

Great post Fred.

The other channel that will have a big play for larger media will be direct self-service ads. Though this does not reach the Coke and Ford's of the world, it does play to the 500,000 other advertisers online today.

[anonymous]

i dont get it... it cant be done under one roof? when your companies get purchased... so begins another cycle of industry consolidation. it only seems natural it will happen here too? or is consolidation a bad thing? it has been my understanding that fragmented markets are the best place to build a business.

[Altay]

How come, if someone clicks on a Google ad in my blog, I get a portion of the clickthrough revenue, but if someone clicks on an ad in my MySpace page, I don't? In both cases, it's my content that's attracted the eyeballs to the ads.

That being said, the content creators on MySpace don't seem to mind.

I wonder how this double-standard is going to play out as the ad revolution continues.

[Beth Gapo]

The enlargement of networks, an ad algorithms, self-service, etc. I agree completely.However, many young players in advertising technology will also want to have abusiness license its technology (eg, ad click.com) to the media and onlinepublishers with large sales teams and advertisers can keep "his" income, and then just reimburse the licensing of expertise.

Free Movies Online