I recently sat through some great presentations on “programmatic direct” media buying at the recent Tech for Direct event in New York. With almost 70% of digital display dollars flowing through the negotiated (RFP) market, everyone wants to be in the game. One of the presenters, John Ramey of iSocket talked about what has happened to the advertising yield curve for digital display. This curve starts at the upper left corner with premium inventory capturing the highest CPMs, and is supposed to flow gently downward on the x-axis, towards the lowest value of inventory, ending on the lower right corner. A classic marketplace yield curve.  In this world, ESPN can charge $20 CPMs for their baseball section, sites like Deadspin in the mid-tail can charge $7, and the networks and exchanges aggregating hundreds of sports blogs in the long tail can charge $1. Nice and fair, and rational.

This is not what has happened, though.

As Ramey correctly points out, we have a yield cliff now. This is world in which there are two types of inventory: The super-premium, which is hand sold directly for double-digit CPMs; and the remnant, which is sold via RTB on exchanges or surviving ad networks, often for pennies. In this world of the Haves and Have-Nots, there is no middle class of inventory—even though one could argue that $7 inventory on Deadspin might actually outperform its upscale cousin, ESPN. This inventory disparity we have created in the digital advertising industry has nothing to do with supply and demand, but everything to do with the process by which we transact.

Premium mid-tail buying is a great idea. Back in 2009, marketplace platforms like TRAFFIQ were bringing this innovation to the space, and enabling marketers to cherry pick and aggregate premium quality sites that could offer friendly CPMs and URL-level transparency. It’s not a new concept. In fact, I think premium mid tail buying is the canary in the coalmine for programmatic direct; when today’s technology can make it easy to put together a large array of guaranteed buys, and enable fast and easy optimization, then we will have succeeded. Here what was missing in 2009, and what we need to succeed today:

  • A Centralized Directory:  You can’t buy stuff without knowing what’s available and how much it costs. Other media channels like direct mail have published prices for mailing lists, right down to audience targeting. You want to reach people who have bought something from the Cabela’s catalog in the last six months, and restrict the mailing to men only? No problem. You can find out what it costs, and who sells it. The digital display market needs to be organized in a directory, down to the placement level. You shouldn’t have to wait for an e-mail back from an RFP to find out what known inventory costs. That work is being done now, but has a lot more work to go through before it is comprehensive.
  • An Extensible Platform: Today’s API-driven technology makes it easy to enable buying directly into publishers’ inventory. A link into DFP means buyers can discover availability and start serving ads with a few button clicks. The problem is that agencies want a Single System to Rule Them All. So far, agencies have been stuck with installed, legacy systems that have more to do with billing and reconciliation than media planning and buying. Agencies want new, web-based ways to discover and buy great inventory, but they also want a system that plugs into their existing tools. They are not going to log into another buying system if they don’t have to. A system that can enable premium mid-tail buying at scale either has to integrate directly into existing media management systems—or replace them. Right now, there are a lot of tech companies at work retrofitting old technology or creating new technology that promises to make this a 2014 reality. It’s a horse race, and agencies are starting to place their bets. The winners are the one with the most extensible platforms that are good at integration, and they will be richly rewarded. The rest will fail, or become a point solution in someone else’s platform.
  • The Right Model: This is may be the most important factor in determining programmatic direct success. If you are charging anywhere north of 10% (and some would argue a LOT less than that) to help media buyers aggregate inventory, then you are not a “programmatic direct” technology company. You are an ad network, or media rep firm. The reason for industry consolidation is because disintermediation through technology has its own yield curve: The disruption that occurs always benefits the middle layer first, but markets always rationalize later. Mike Leo, former Operative CEO, told me about how another industry solved a similar problem that was occurring in the media business, where ad agencies were starting to rebel against specialized media buyers who in the middle of the transaction, with opaque pricing methodologies. The year was 1968, and agencies teamed up and decided that a standard rate of 15% was all they were willing to pay for television buying services (and then they eventually bought all of the media buying companies, but that’s another story). Anyway, markets always rationalize themselves, and right now even 15% feels like a big vigorish for agencies with ever-shrinking margins on their media practice.
  • Standards: It’s 2013, and we are still faxing IOs. This is largely because there are no accepted standards—and no protocol—for electronic orders. This is actually not a hard problem to solve, but getting adoption from buyers and sellers is what’s needed. Right now, a few companies are working with groups like the IAB to get real traction with standards, and we need that to succeed to make programmatic direct buying a scalable reality. Electronic orders suck a lot of the viscosity out of the deal pipeline, and start to let the machines do the grunt work of order processing, rather than a $50,000 junior media planner.

The good news is that there has been a tremendous amount of progress in 2013 on all of these initiatives. The promise of true programmatic direct buying is closer than ever, and there is enough real development behind the hype to make these dreams of efficient media buying a reality in the near future. In that future, it just may be possible for a buyer to use demand-side technology to aggregate the “fat middle” of premium mid tail publishers, and start to reward the middle class of inventory owners who are currently getting paid beer prices for champagne content.