Commentary

CPMs Headed Up?

  • by , Featured Contributor, January 24, 2008
Are online media CPMs headed up or down? The question is on a lot of people's minds these days. JPMorgan analysts, in their "2008 Internet Investment Guide, Nothing but Net" which was released last month, expects the CPMs of graphical online ads to grow 3.9% and the number of ads per page view to increase 8.3%. Overall, they expect revenue per page (RPM) to grow 12.6% in 2008.

Wow! Sounds like a great time to be an online publisher. What do they think is going to drive that growth? They cited four primary drivers: lower comparables from 2007, when ad rates softened from the strong growth of online ad networks selling nonpremium inventory; better inventory sell-through; behavioral and geographic targeting; and ad exchanges. While this sounds logical and may prove to be true, I don't think that it tells the whole story.

We will certainly see ad rates go up this year for certain types of inventory, but I suspect that we are also going to see ad rates go down for a lot of other types of inventory.

advertisement

advertisement

Who will be the winners? I think that ad rates are going to go up significantly for vertical inventory on premium branded content sites. The shift of more and more brand dollars into online will drive that all by itself, plus we are starting to see much better inventory yield management, sell-throughs and the utilization of targeting on these sites. I also expect to see CPM growth on social networks. It won't be hard, since these networks are driving off of such a low base as it is, but they will certainly benefit from much better targeting, whether behavioral or preference-driven, and from the growing acceptability of their environment among brand advertisers. Finally, I think that we are going to see CPM growth on small and niche content sites, as ad networks from FM Publishing, to Google AdSense, to the ad platforms and exchanges that the portals are creating, all do a better job providing marketers with targeted access to these sites.

OK, that part was pretty easy, but who will be the losers?

I think that we will see ad rates go down for virtually all undifferentiated graphical ad inventory, particularly on big sites and portals. Why? Because the shift to ad networks, sophisticated forms of audience targeting and performance optimization has commoditized this inventory. It used to be that these sites could get premiums for just delivering tonnage, but no more. Ad networks can deliver much more tonnage, and do it with cheaper underlying inventory. Further, and maybe more important, much of the value that the ad networks create, they get to keep. The value in large blocks of online media is shifting from the page to the person. The companies that control the audience data, insights, targeting systems and sales channels are creating and keeping that value. More and more of this inventory moves in arbitrage-driven deals, meaning that ad networks are no longer the low-margin businesses that ad representation firms have historically been in other media. Since they typically operate on low incremental cost structures, they can drive very significant profits in inventory with low CPMs. Not so for most content companies.

What does this mean for folks in the online ad ecosystem? I think that publishers and media sellers needs to be very cognizant that hearing good news about the growth in "blended" online ad rates is great, but may mask serious weaknesses that some of them have in core parts of their business. For ad buyers, the rise in rates for premium vertical content sites will probably mean the need to explore ad network and social network opportunities much more, since clients are focusing more and more on ROI in their buys. For everyone, it certainly means spending more time and attention understanding how the value is flowing from pages to people, and making the investments in analytics and targeting systems to try to capture and control as much of that value shift as possible. What do you think?

Next story loading loading..