Online advertising landscape

Lately I have been paying a lot of attention to certain dynamics in online advertising and been giving some thought to how they will eventually play out. For this post I am specifically referring to the interactions between 3 types of companies:

1) Ad networks (e.g., Google AdSense and many others)
2) Ad exchanges (e.g., ADSDAQ and a few others)
3) Intermediaries/optimizers (e.g., Rubicon Project, Pubmatic, AdReady and a few others)

The main goal of the first two is the same: bring together advertisers and publishers. Both are used by publishers for their remnant inventory (stuff they can't sell directly) or by small publishers for all their inventory. But today there are differences between them.

First, there are different types of ad networks, mainly in terms of:

- type of ads they serve (e.g., display, text)
- audience they are focused on (e.g., women, hispanics)
- how they target the ads (e.g., contextually, behaviorally)
- geography where they target the ads (e.g., North America, Latin America)

Ad exchanges on the surface are fairly similar to each other, but they haven't been around as long.

So, what are the differences, and what is the interplay, between ad networks and ad exchanges?

Both want to bring together advertisers and publishers and both want to maximize the value created in the transaction. But they have philosophical differences. Exchanges are using financial markets as models (evidenced by some of their names, e.g., ADSDAQ, AdECN), with auction models for matching and full transaction transparency. And they take a small % of the transaction as a commission. Ad networks, on the other hand, are more like black boxes, both in terms of how a match is made and what % of revenue is paid out.

In theory then, the exchange model should be superior to the network model (for everyone but the network). And you would expect networks to be under pressure to become more open and transparent, eventually looking more and more like exchanges themselves. Taking it a step further, there really should only be one big exchange - exchanges benefit from scale and, like any other platform with network effects, the more people on an exchange the more value is created. Many people see this and that is why Microsoft bought AdECN, Yahoo bought Right Media and Google now has the Doubleclick Exchange. They are hedging their bets and preparing for what will come eventually.

However, in the short/medium term this is unlikely to happen (especially the one big exchange part). First of all, there is inertia. Second, ad networks (especially Google) will fight to keep their current model for as long as possible. Third, the online marketplace is still young and ad networks that focus on specific niches do add some value and provide access to publishers/advertisers not available elsewhere. This will change, but it could take a little while.

For reference, look at how long financial markets have been around, how long it took for things to move to predominantly electronic exchanges, and the fact that it has only been in the last few years that consolidation has started in earnest. The advertising market will follow a similar, although accelerated trajectory, but we are still in the early stages.

Great, so you can make a case that all ad networks will morph into exchanges, and eventually you may end up with one huge exchange (or, more likely, 3-4 fairly huge ones).

So what about the third category (intermediaries/optimizers) I started off with and what is their role in all this?

I'll leave that for the next post because this one is already long enough...