Imagine an estate auction with 100 interested bidders and myriad different pieces on the docket, ranging from the small and insignificant to larger and historically relevant. Because each of the potential buyers has perfect information and an equal chance to bid on any piece that they find valuable, once the bidding begins, the efficiencies that result from the open auction model will take hold, and this time-tested execution will produce the highest price possible for each item. If, however, we were to separate out the 100 potential buyers into 10 groups of 10, and only show each group a small unique set of the total pieces, the over all market efficiency (defined as maximum monetization) of this model would fall. The very strength of the open auction model depends on transparency and equal access.
So what does this have to do with programmatic media buying and how publishers monetize their inventory?
From the early days of online advertising, many publishers have organized potential buyers of their inventory into sets, leveraging a "waterfall" model. Here, publishers sequentially line up partners based on historical performance or known expectations. They offer unique buckets of inventory to distinct monetization partners that will then monetize the impressions they are allocated to the highest paying buyer within their own platform or exchange.
If the monetization partner cannot fill the impression above a reserve price, they will typically pass the impression back to the publisher. Even when the process of progressing down the waterfall is automated, each impression is still going out to one potential monetization partner and then coming back to the publisher's ad server, only to be automatically sent out again to another monetization partner until the impression is filled.
This waterfall of monetization opportunities typically involves a publisher's direct "reserved" campaigns being placed at the top of the rotation with a number of supply-side platforms or other types of network partners being prioritized thereafter in the chain based on price and fill rate. While value can be derived through this process, there are significant inefficiencies left on the table, as impressions are often lost in each transaction, especially when passed from one monetization partner to another down the chain. In an ideal world where the process could actually live up to the real-time bidding moniker, multiple monetization partners would compete against one another within the ad server for the same impression at the same time, with the highest bidder winning the impression.
There are, in fact, growing trends in publisher monetization that could both see the end of the waterfall and deliver the promise of true programmatic efficiency to the process of inventory monetization. Dynamic auction technology has now evolved to the point where, by allowing one dedicated exchange partner to take sole responsibility for the monetization automation, a publisher's inventory can be auctioned in real time, with every impression offered simultaneously to every interested bidder, at scale. Further, more publishers are becoming comfortable in allowing their trusted technology providers to place proprietary code on their sites that allow the auctions to occur and be optimized inside the ad server. This allows multiple monetization partners to pass in a bid on the same impression, dramatically increasing the competition for each single impression served.
Programmatic media buying only continues to grow in importance, and publishers have rushed to realize the value it presents. This has led to executions that have sometimes felt forced rather than organic. The waterfall, itself, represents a vestige of a pre-automation supply-side model. Only by accepting a shifting paradigm and embracing new models that maximize the advances in technology will publishers realize the benefits of a truly programmatic landscape.
On Twitter? Follow iMedia at @iMediaTweet.