• by Matthew Greitzer , Op-Ed Contributor, July 17, 2009

I've written several times over the last year about the emerging opportunity in auction-based display media, especially as it relates to ad exchanges. The ad exchange landscape is a close cousin to search marketing, borrowing its auction-based pricing model, real-time spot markets for inventory, and the application of micro-segmentation to drive results.

A small group of us at my company hypothesized that these "search-like" capabilities applied to display media could produce "search-like" performance, even though display media is a push, not pull, medium. So far our hypothesis has been correct. And though the ad exchange market is a tiny fraction of the $30 billion global search market, we think it's a channel worth betting on.

Apparently Google does, too. The company's on-again, off-again, enthusiasm for ad exchanges is on again, and this time for real. In the next few weeks Google will likely release version 2.0 of its DoubleClick ad exchange, offering several major upgrades from its predecessor, the most notable of which are impression-level bidding and the on-boarding of Google Content Network inventory.

Impression-level bidding is the pinnacle of ad exchange sophistication. It enables visibility into 100% of available ad impressions running across the exchange at any given point in time, along with data points about the users viewing those impressions, such as connection speed, geography, and other cookie-related targeting information. There is no applicable analogy to this in search. Even with all the exposed targeting parameters available in SEM (keyword, geography, time of day) it doesn't come close to the potential for micro-targeting and pricing optimization enabled by impression-level bidding. For marketers who champion data-driven media buying decisions, impression-level bidding is their paradise.

One of the reasons I have focused on exchanges and the opportunities therein is because I believe they are the next frontier for the search marketing community. Google, however, does not agree. At least that is how I interpret its gating requirements for DoubleClick ad exchange participants. At present, the DoubleClick exchange is open only to ad networks (or network-like entities. My company can participate because it has a small team called ATOM Systems that aggregates buy-side demand and operates in a quasi-network model). But most search marketers, whether in-house or served by agencies, will not have direct access to the exchange, and will miss out on its inherent benefits.

On one hand this seems an odd choice by Google. Its strategists have built their business, one of the most successful in history, by aggregating demand through easy, open access to high quality ad inventory. And here we find them cutting off a huge swath of demand by excluding direct buyers, and agencies buying direct on behalf of clients.

On the other hand, it's an obvious choice. Google has to convince its high-quality publishers using the DoubleClick ad server to adopt the exchange to liquidate unsold inventory. These publishers will be reluctant to play in this space if they feel they are undercutting their direct sales force by offering the same inventory for less in a spot market. So direct buyers, and their representatives, are out! They will need to buy from publisher sales reps directly, or from Google's own sales force -- which is shaping up, more and more, like an ad network.

I understand Google's rationale, and I think we can assume this is the first step in a many-stage evolution. Perhaps six months from now Google will open the exchange to all comers. Or maybe that expectation is too optimistic. If not that, then perhaps the company will enable the targeting and pricing benefits of impression-level bidding to the Google Content Network alone. This would be a benefit to Google, its GCN publishers, and advertisers alike, as the improved targeting and pricing capabilities would create higher ROI, driving more dollars through the GCN network. And cannibalization of direct sales, at least for the long tail of the GCN, is a non-issue.

Here's hoping Google moves in the direction of continued, open access to its advertising platforms and services. In the meantime, keep an eye on this nascent but rapidly developing industry segment.