Digital Marketing Data and Economies of Scale

The recent Omnicom-Publicis merger talks have sparked an intriguing debate: does one’s scale of data and transactions translate into a buying power advantage in digital media? This was certainly the initial take offered by many, such as David Bank, an analyst with RBC Capital Markets who was quoted in the Wall Street Journal article announcing the merger:

“The bigger an ad agency is, the more likely they are to have more access to consumer data and data around the pricing of online ad impressions,” which helps them get the best price when buying from “digital behemoths like Google and Facebook.”

Others, such as Dave Morgan of Simulmedia, disagree (as quoted by Peter Kafka in All Things D):

“I think it’s a total misdirection to think that you can leverage the scale and advantages of big data if you’re bigger. Quite the opposite…Most of [Google et. al.’s] media is auction-based. The fact that you buy more doesn’t let you get a better price.”

So, which is it? Do more data and buying scale translate into a buying power advantage in auction-based digital media or not?

Of course, the answer is not so simple. First, let’s acknowledge that Dave Morgan has a point: the auction systems employed by Google and others are cleverly designed to encourage bidders to issue bids as close as possible to the “true value” they assign to the media they seek: both overbidding and underbidding are suboptimal strategies in a Google-style auction. Buying volume seems to offer no direct leverage when impressions are bid on one-at-a-time. This is one of the wonders of real-time bidding: in principle, it maximizes value for all parties, regardless of scale.

So are there really no advantages to scale in programmatic media? Putting collusion aside, the more historical information you have about marketplace conditions based on aggregate observations, the better you can predict both what an impression’s true value will be relative to underlying marketing objectives, and where relevant market prices are likely headed, leading to superior bidding strategies. This kind of advantage comes not just from data, but also from technology that enables one to capture and analyze large volumes of trading information in near real-time. Such techniques may not necessarily lower prices, but they will result in better value.

There’s also the obvious observation that not all digital media is sold at auction: much more revenue still flows through direct channels, where data and volume continue to deliver negotiating leverage. As approaches like programmatic premium continue to make inroads, these advantages will continue to grow in significance.

So, in conclusion, when it comes to data-driven media operations, size does matter. Naturally, so does expertise. Consider both when evaluating options for data-driven marketing.