Real-Time With SiteScout's Matt Sauls On Programmatic Advertising

Programmatic ad buying is slowly being used for more "premium" ad buys, but what does "premium" mean? In fact, what can "programmatic" mean in this case?

RTM Daily spoke with Matt Sauls, SiteScout's VP of operations, to see where the industry stands on programmatic buying, premium inventory, and whether or not there is any potential pitfall to putting too much stock in viewability ratings. SiteScout was recently acquired by Centro.

RTM Daily:
We are going to be talking about "premium" and "programmatic," so before we even begin we need to define both. What does each term mean?

Matt Sauls: Programmatic can mean any technology that automates digital media workflow. It does not necessarily eliminate human contact in the sales process (although there are times when it will), but it does reduce human labor in the execution process. It eliminates the human involvement from most of the middle steps (I/Os, building tags, delivery reports, etc.). 

Specifically, programmatic can be technology that allows a buyer to input simple targeting requirements (budget, geo, context, etc.) and hit the proverbial “Buy Now” button. In this case, the buy instructions will be automatically accepted/rejected by the publisher’s ad server based on avails. If accepted, the buy is inserted as a line item into the publisher’s ad server. This is ‘programmatic direct.’

In more complex forms of programmatic media, buyers use demand-side platforms to target inventory that has been made available by publishers via RTB. Buyers can set complex targeting instructions, filter impressions using various data, indicate optimization settings and gain access to massive amounts of inventory through mostly automated means.

"Premium" is difficult to define, as it is more open to interpretation. In the earlier days of RTB (2009-2011), most available inventory was "remnant" -- inventory that publishers couldn’t sell and was not generating revenue. This would usually be less attractive inventory, possibly appearing at a higher session depth, in less popular geos, or on less desirable sections of a site. In programmatic, "premium" has generally been used to describe inventory which is not remnant. As the RTB ecosystem matures, publishers are making much more of their inventory available, and RTB buyers can now access a lot of inventory that would previously have been sold direct. The term "premium" can mean that a certain inventory is more desirable, performs better, or demands a higher CPM. For the use here, we should think of premium as the opposite of remnant, and not as a description of the quality of certain inventory.

RTMD: The ad industry seems to be making a push toward "premium" programmatic. Is this push manifesting most through real-time bidding, programmatic direct offerings, or private exchanges (which can be a mix of RTB and direct)?

Sauls: Premium programmatic is manifesting through the mix of RTB and private exchanges.

Although “programmatic direct” was the first way in which the industry saw non-remnant inventory becoming programmatically available, it seems unlikely that this will be the driver of the trend. At this time, publishers are eager to see their entire inventory compete in an equal arena, where direct buyers compete with bidders to drive CPM prices higher. Simultaneously, media buyers are constantly looking for efficiencies and need critical mass of inventory to be accessible through a single platform. Right now, there is no critical mass. This means that some form of RTB-based buying will be the solution.

Private exchanges are currently the biggest source of premium programmatic. A number of "premium’" private exchanges are only open to select buyers and see much higher average CPMs. These exchanges have inventory which the publishers wouldn’t offer on the open market, but they operate like standard RTB, making things easy.

Private Marketplace (PMP) or DealID is another promising development. Realistically, however, the industry still lacks automated communication methods for negotiating PMP deals, causing them to be inefficient and negating the real value of programmatic buying. Although PMP/DealID transactions should have a very bright future, they are not as widespread as some exchanges would like people to believe. More automation in PMP could change this significantly. This is what many exchanges are counting on. 

RTMD: What percentage of inventory sold through open exchanges would you say currently qualifies as "premium"? 

Sauls: We don’t have enough information to answer that question, and I don’t think any single tech vendor in the space can honestly make that estimate. This is because many of the most legitimate premium programmatic offerings are still happening privately and are hard to measure.

RTMD: Do you think that number will increase, or have open exchanges found a home in selling remnant inventory?

Sauls: Yes, the proliferation of Private Marketplace (DealID) buying should greatly increase the amount of premium inventory sold and private transactions conducted through open marketplaces. 

RTMD: Now on to private exchanges -- are they really all that more "premium," or does it simply give publishers more control and make advertisers feel like they are a part of something more elite than open exchanges?

Sauls: Yes, there are private exchanges that truly sit much higher in publishers’ waterfalls. These exchanges tend to have higher CPM prices and enjoy inventory that is at a lower session depth and in more attractive locations on publishers’ Web sites. These exchanges tend to be lesser known, as they are usually very picky about who is allowed access.

RTMD: Viewability as a metric has been given the thumbs up. What role does viewability play in making ad inventory "premium" or not? 

Sauls: This brings up the dichotomy of meanings discussed in the first question. While viewability is certainly a measure of inventory quality (and an important measurement), it is not necessarily what is meant by "premium," in that premium is often used to mean the opposite of remnant (inventory which the publisher was unable to sell directly). In the future, when publishers make most/all of their inventory available programmatically, it should make more sense to begin using the phrase ‘premium’ to refer to the quality of the inventory more specifically. For now, let’s say that yes, “viewability plays an important role in making inventory high quality,” and that a premium exchange would be expected to have high(er) quality. 

RTMD: Are there any dangers associated with holding viewability metrics in such high regard? What are they? 

Sauls: One danger is the mistake of measuring viewability on a per-site, per-publisher or per-placement basis, in aggregate -- instead of on a per-impression basis. In this case, publishers could see large chunks of inventory rendered "low quality" when in reality each time a user views a page, something different can happen. That said, with modern programmatic systems, measurements are being taken more and more accurately, so this should not be the case. Furthermore, if publishers are seeing inventory that is viewable only a small percentage of the time, it stands to reason that although this may hurt their bottom line, advertisers should not be paying for it in all fairness. 

RTMD: Is programmatic's future more in automating direct deals -- which would be a blend of past and present techniques -- or is it more futuristic, meaning automation from start to finish? 

Sauls: In the near future, automation of direct deals is plausible. While many platforms do have amazingly hands-off automation, there is still a ton of human intelligence that goes into planning high-impact brand campaigns. Especially for long sales-cycle products, CPGs, etc., where performance metrics are less useful, viewability and access to premium, high-quality inventory is key. Advertisers will always want to guarantee certain placements, in certain volumes. Programmatic technology is unlikely to replace the process for executing custom or high-impact deals such as wallpapers and takeovers.

Increasingly, we expect to see savvy advertisers understanding the benefits to buying on a non-guaranteed basis. In the past, they could not secure high-quality inventory in this way, but with dynamic allocation, deal IDs, and the transparency of the RTB market, advertisers can choose across a broad spread of high-quality sources. There should be some price efficiencies for the advertiser as they forego guarantees of delivery. Over enough inventory sources, there should always be enough impressions to fill campaign goals. Publishers can also expect that advertisers who buy in this way will pay much more than the current going rates in the wide-open RTB market. Private Marketplace/Deal ID is certainly poised to offer the most immediate improvement in workflow efficiency while allowing all of the advanced targeting and optimization that RTB platforms offer. 

Realistically, "automation from start to finish" is still a less plausible concept in the brand marketing world and is much more possible in the performance marketing world.

Recommend (2)