As Saul Hansell points out, several companies at ad:tech last week were talking about their new vertically focused/behaviorally-targeted/blah blah new ad networks. There are more advertising networks than ever, as well as ad network enablers (”start your own ad network” says Pulse360, Adify has been saying this for a while) not to mention advertising exchanges and marketplaces (Right Media, Traffiq, Adsdaq (Contextweb), AdECN (Microsoft), Doubleclick’s, AdBidCentral), publisher optimization tools that are de facto marketplaces (The Rubicon Project, Pubmatic), open source ad servers with apparent scaling problems that want to be open source ad networks (OpenAds) and so on. Disclaimer: we’ve launched two ad networks, and spent a lot of time designing ad servers, behavioral targeting systems and other fun things like that.

The simple reason why ad networks are so attractive is the same reason that comparison shopping sites are much more interesting than most online retailers except the first and best, Amazon.com, ever have been: ability to scale and resulting increasing returns. It sounds trite, but companies have been learning that instead of confining their efforts to selling content on sites that they own and operate, the same adserving infrastructure, ad sales people, media planning and tracking infrastructure (by the way, much of this on the publisher and even network side is still woefully inadequate from a workflow and efficiency perspective) can be leveraged across tens and hundreds of websites. The ability to roll up and create consumer segments and enhanced targeting products is also quite a juicy and appealing notion. The former is more of an advantage (albeit short-lived) than the latter, the latter being perceived as much more of a factor than it really ends up being.

We see some things changing here: some vertical ad networks are going to disappear altogether or morph into barebones data collectors and providers of cookie-based data to larger networks; larger non-PII data aggregators are going to emerge, along the lines of eXelate which has a very nice model; publisher tool providers like Pubmatic will make some inroads, but in the medium term be unsuccessful in displacing Google Adsense as the default choice of the long tail; free ad serving will become a factor for small and mid-sized publishers but although it has a large lead, OpenAds will not be the big winner in this area due to technical problems - Google will be a factor but somehow will end up shooting themselves in the foot over this issue SO the main way the free adserving has an impact is in the increasing use of geotargeting; thus leading us to see that international ad sales organizations / agencies which may or may not be functionally equivalent to international ad networks, will actually do well among all of these.

Whew, just a few of our thoughts on ad networks– we have plenty more to put forth on this topic though to summarize, we think a lot of the generic ones will fall by the wayside in the next 18 months and some will morph into something else. Most interesting will be that small but select group that will grow and add value through scale, international CPMs, unique targeting (we remain skeptical), sheer speed, or workflow/sales/creative innovation.