• by Mark Simon , May 21, 2007

Microsoft may have lost the bidding war on DoubleClick, but it's clearly shown no aversion to paying up to grow its new-media empire. On Friday, the giant from Redmond promised $6B in cash for aQuantive, Inc. -- nearly double the $3.1B that Google will shell out for DoubleClick. Message to Google: don't rule Microsoft out.

But while there's a lot to say about how Microsoft remains a serious contender for the online throne, I want to focus on what Microsoft/aQuantive -- and the surrounding flurry of networks that are buying up ad agencies -- means for advertisers. And the message to advertisers is clear: the agency/network divide is starting to blur, and that might not be a good thing.

To be sure, network/agency marriages-from Microsoft/aQuantive, to Google/DoubleClick, to Yahoo/Right Media -- aren't exclusive to the online world. Earlier this year, media giant Meredith Corporation -- owner of magazine titles including Ladies Home Journal and Better Homes and Gardens -- purchased word-of-mouth agency New Media Strategies. Meredith also bought digital agency Genex and, in 2006, picked up interactive agency O'Grady Meyers. In a January article about Meredith's acquisitions, The Wall Street Journal's Emily Steel also noted that Conde Nast and Wenner Media (owner of Rolling Stone) have both recently established in-house ad shops.

But it's in the online world where the new conglomeration seems to be focused. And it's in the online world, as well, where the new conglomeration has the potential to cause real trouble.

Why? First, ironically, because the online world offers richer analytics -- and that gives more opportunity for any network-agency conflicts of interest to cause problems

On the one hand, analytics should shield advertisers from any conflict of interest: if agencies steer advertisers to buy too much of a parent network's inventory, the ROI numbers will show that something's amiss. More data is the advertiser's best protection.

But if richer online analytics mean more transparency, they also mean more complicated media buys than ever before. Even if advertisers can more clearly spot an ad that's getting poor ROI, it's also much easier to hide lost opportunity in a maze of data -- especially in the fast pace of real-time auctions. Online advertising is highly transparent -- but it's also highly confusing, and so the need for a third-party advocate doesn't go away in the online world. Unfortunately, network/agency conglomeration makes that kind of advocacy much harder to find.

In the end, though, while the potential for abuse is huge, the networks also value their advertisers' trust -- which is why I'm not highly concerned about Microsoft, Yahoo, or Google using their recently acquired agencies to lead advertisers astray. What the networks do have an incentive to do, however, is to push competitors' agency clients away from their APIs. Which is my second reason for saying that, in the online world in particular, network/agency consolidation can be bad for advertisers.

Let me explain. Running a campaign through an ad network is an ideal way to learn how it operates. It teaches you how each network performs for each type of client and each type of vertical. Which is why the networks could use their new agencies to gather high-level intelligence on their competitors. Google could use DoubleClick's Performics to study MSN's AdCenter. AQuantive's Atlas could be used by Microsoft to gain intelligence about Google.

Of course, the networks already have some experience with the others' platforms -- Microsoft, for instance, already runs ads within Google for Microsoft products. But for now, Microsoft's intelligence about Google is limited to what it's learned through its own campaign. If Microsoft gains access to all the data, across all the engines, for aQuantive's entire client roster of search clients, it will be sitting on a treasure trove of information that it's never seen before -- and which should have Google feeling very nervous. The same is true, of course, for the information that DoubleClick's Performics can provide to Google. To a network, an agency is a wealth of competitive data -- a fact about which all of the networks are undoubtedly aware.

Even if none of the networks have any intention of using their agencies as espionage tools, the threat is still real. And since the threat is real, the networks will be forced to consider pushing competitors' agencies out from their APIs. And if the agencies do act on their fears, their clients -- that is, the advertisers -- will become collateral damage in the network wars.

And so it remains to be seen whether the new conglomeration of agencies and networks is good for advertisers, or if it's a new negative force in the advertising landscape. Which is why my advice to the networks is that they think, first and foremost, of how they can keep the agency/network silos firmly in place, or as firmly in place as possible.

If they can't keep those silos in place, then we may see an advertiser revolt in the future.