Why now? In other words, why has this programmatic direct trend been on the radar lately? What’s driving all the conversation?

It’s really something my boss Joe Pych calls the “Sutton Pivot”, inspired by the famous thief Willie Sutton who robbed banks “because that’s where the money was”. Over 70% of digital display dollars are transacted in a very manual way today.

Despite all the LUMAscape hype over real-time bidding (RTB), most of the digital money still gets transacted through the request-for-proposal (RFP) process.

Everybody wants a piece of the action, hence the “Sutton Pivot”, in which all the ad tech companies are running to try and provide automation technology for directly sold deals. It’s actually a good thing. Today’s process for buying guaranteed digital media can take over 40 steps and suck up over 10% of media budgets just in man hours.

The concept of ‘programmatic direct’ or ‘programmatic premium’ is a relatively new phenomenon, but it’s really just about automating the buying process for digital media, right? What makes it different from the automation happening in real-time bidding? What’s the difference?

Real-time bidding, or what we are starting to call ‘programmatic RTB’ has been a real boon to the industry. We now have a set of ‘pipes’ which connect demand- and supply-side platforms that make the digital media procurement process hugely efficient.

Today’s systems are modern, cloud-based, scalable, and have super low latency. We are seeing the type of liquidity and deal flow that happens in systems like NYSE and NASDAQ.

That said, 70% of buying that happens in digital is neither ‘real-time’ nor ‘bidded’. It’s just two organisations trying to make a deal. You need different technology to enable that kind of guaranteed transaction, and marketers are starting to wonder why they are paying so much in transactional costs to access higher classes of digital inventory.

RTB proved that efficiency can happen in digital, and now marketers want faster and more efficient access to more than just remnant inventory.

You say that agencies have a ‘perverse incentive’ to embrace efficiency in buying. It would seem counter to everything that is happening in the programmatic space at the moment. How do demand-side business models need to adapt for programmatic direct to become a reality?

Agencies make money when plans take 400 hours to create. Manually trafficking line items in an ad server, and cutting and pasting publisher insertion orders pays the bills for agencies who charge on a cost-plus basis.

Digital media agencies have been operating that way for years: hire cheap, work the ’23-year-old media planner’ hard, and earn a mark-up on their labor. Nothing wrong with work-for hire, but the RTB phenomenon – and what marketers experience with easy-to-use programmatic platforms in search and social marketing – have changed the dynamic entirely. Agencies have to do more than heavy lifting now to survive.

They need to hire fewer, smarter, people to leverage systems – and more great creative and analytical people to make sure they are driving digital messages that inspire and meet KPIs. The days of getting paid to traffic ads in MediaVisor are over. That’s a big time cultural change for agencies. A lot of shops won’t survive the transition, and that’s a good thing.

What are some of the things – beyond cultural change – that need to happen to create this new era of programmatic direct efficiency? What’s missing?

We tend to think of digital as this highly advanced form of marketing, but it’s really the most backwards. Direct mail costs something like $750 per thousand (CPM) to put a catalog in the mail – and marketers like L.L.Bean make that number work consistently.

Digital struggles to make $10 CPA goals work on $5 products. That’s really lame. Part of the problem is the lack of basic information available to the marketer. If I want to buy a direct mail list, I can find out how many folks in the list live in San Francisco, and have purchased a product by credit card in the last month. I can find out how much it costs to by that list – and who sells it.

Until recently digital media has had no such directory. Not only that, but the industry lacks even the most basic set of electronic ordering protocols, that can enable systems to understand each other in electronic transactions.

The good news is that more work has occurred on this front in the last two weeks than has happened in the five years the IAB has been promoting ‘eBusiness’ initiatives. Look for some significant announcements in this area soon.

Who benefits most from adopting programmatic direct strategies? Publishers? Agencies? The marketers themselves? Are there winners and losers if this new tactic sees adoption at scale?

It’s easy to say that everyone’s a winner with programmatic direct adoption at scale, but that’s not entirely true. I think publishers are the big winners, because they are starting to take some control back over the procurement process from the demand side.

I think longer tail sites that depend on RTB revenue streams will continue to be able to get access to demand at scale through RTB systems, and still get their AdSense money.

But what really excites me is seeing high-quality publishers that own high-quality real estate on category-specific properties finally get more control over pricing and partner selection. This will be even more critical as publishers expand their offerings cross-channel, into video.

Publishers need a programmatic way to sell their higher classes of inventory, and not be so dependent on prevailing procurement methodologies which overvalues biddable, commoditised inventory. Agencies who value higher class inventory also win, of course.

Right now, the conversion (and action) seems limited to display media. How does programmatic direct impact cross-channel buying?

Everything digital will be bought programmatically in five years. Some will be RTB display, and some will be display, native, and video inventory purchased through programmatic direct platforms. Addressable television, digital out-of-home (DOOH), and other channels will also factor in.

Once we can get a true unique identifier that makes sense from a technology and privacy standpoint (big question, obviously), then marketers will really be living in programmatic heaven.

You’ve been working in the programmatic direct space for a long time (starting at TRAFFIQ in 2008), and yet there seems to be fairly little adoption of the concept among agencies. Are you crazy? Why keep doing it? Will there be a big payoff in the end?

Change is really hard, especially when the pace of change is as rapid as in digital ad technology. When I was on the publisher sales side, there was always something that bothered me about getting a $200,000 insertion order for digital advertising through a fax machine. That stuff still happens today.

Ultimately, I so believe that true process automation will happen in digital media, and that we can free people in the space to stop doing a lot of manual grunt work, and start being truly creative.

I was watching a documentary the other night, and an engineer was talking about why he loved his job. He said he spent the last three years building a bridge that eliminated 10 minutes from the commute for some 20,000 people a day. “I saved people over 50,000 days of productivity last year,” the engineer explained, adding, “I wonder what those people are doing with all that extra time”.

There are a lot of young people who go into an agency thinking that they are going to help make the next kick-ass viral ad, but they end up working until 10 o’clock at night pasting line items into an ad server. I really think that, if we can change that, great things will happen.