“Digital Is Dead” read the first slide of Larry Allen’s keynote for Programmania on June 5, 2017.
I don’t know if it elicited the gasps Larry was hoping for, but I could detect a certain sense of puzzlement flowing throughout the room of 200-plus programmatic specialists. Wait—isn’t this supposed to be a conference about digital advertising? Isn’t Larry Allen VP of Programmatic?
AND Ad Innovation. Yet he was kicking off a day of programmatic discussion talking about innovations in linear. Is digital truly dead? No, but the idea was it’s a mess—clogged with too many intermediaries, a lack of transparency, and an abundance of fraud for starters.
The programmatic space has long desired to grab brand TV budgets by extending pipes into the boob tube, but if that’s going to work, our digital model needs to transform drastically. There’s a lot to learn from the simplicity of TV transactions and executions, as well as progress like Open AP (featuring Turner and its buddies Viacom and Fox Networks), which is enhancing linear buying with digital-style targeting capabilities.
It was a strange way to kick off our first ever Programmania event, yet I think it sent the right message: this day of presentations and discussions was trying to push the programmatic conversation out of its comfort zone. I thank all the people who made the inaugural iteration of Programmania a big success, and while I already have some ideas for next year, I wanted to share several points that got my wheels spinning.
1. Let’s start with the publisher red meat: I came up with the title Team S2S vs. Team HB for the server-to-server/header bidding debate as a joke on “Twilight” and the whole bizarre Team Glowing Emo Vamp vs. Team Handsome But Feral Werewolf Guy. At the same time, it does feel like pubs are taking sides with these technologies when hybrid solutions are probably the best approach.
Mike Hannon gave us great insight into the development of Purch’s massive S2S framework that is tied directly into the publisher’s ad server. Oh boy, S2S is fast and efficient, and Purch is racking up the integrations to SSPs and DSPs warp speed. But the ID-syncing issues are real, and Hannon suggested the best way to offset a drop in bid density and/or yield was scale—more (a lot more) hookups.
When it comes to DFP, users are pretty much stuck connecting their ad server to EBDA, which appears to be a black box. Header-based S2S efforts from tech are still ramping up, but whether demand sources will play nice together in these environment and what data comes back through the header is unclear. There were hopes the ID-syncing issues could be cleared up by next year, but still—I wouldn’t kick your “traditional” HB hookups to the curb yet.
2. Enter #Pubtech. Another way to easily scale up your S2S efforts would be to license Purch’s framework for your own site—a tool to enable unified auctions. In fact, all of the representatives on the HB/S2S panel (which also included Washington Post and SheKnows) boasted development teams that had seriously beefed up their back ends, and now were looking to license the tech. A pubtech offering is especially interesting considering that demand sources don’t want to sit within each other’s server-side wrappers for a variety of reasons.
3. It seems PMPs (at least display ones) have made a lot of headway over the last year as agencies and publishers are increasingly communicating and sharing objectives and optimization strategies with their counterparts. Yes, PMPs are being treated like direct sales that require even more hand-holding.
The really interesting thing is that some pubs and agencies are pretty much diagramming their tech stacks for each other, including all the SSPs and DSPs they use. At some point both the buy and sell sides could leverage the obscuring of partners for a bit of advantage, but I think it’s becoming clear that it’s mainly the intermediaries that benefit from a lack of transparency.
4. Let’s play, “What does that ad tech buzzword mean?” Programmatic guaranteed gets particularly tricky—are we talking about using automated channels to buy inventory similar to direct guaranteed buys? Or spending guarantees for PMP campaigns? It’s definitely hodge-podge at the moment, but the end goal seems to be what we used to call audience futures: campaigns guaranteed against reaching a certain amount of an audience segment that are executed through programmatic channels. Doesn’t sound the least bit convoluted…
5. Do agencies and publishers need CMPs? A creative management platform could automate much of the process of building data-driven creative for standard units as well as testing and optimizing. Replace creative with content and you can see how such a tool could work for publisher site-personalization efforts. You would think major martech companies would offer such services, and some build-it-yourself publishers have or are developing similar technology, but something tells me a CMP “point solution” could go a long way in the near term. Eventually, I see CMPs getting assimilated into larger offerings.
6. The rise of people-based or identity marketing has made a lot of publishers without highly vaunted registration data (that hot deterministic stuff) feeling left out. Never fear—agencies like Merkle will partner up with you to figure out the identity bit. But won’t that mean publishers are commoditizing their oh-so-precious first-party data? Possibly…
7. Sure, broadcasters and major digital publishers can easily sell out of all their video inventory directly, but they are probably leaving money on the table if they don’t examine what they could get for that inventory in programmatic video markets.
The question, though: is that potential revenue worth the effort and cost of managing private marketplaces (the open video marketplaces are still too scary for many premium video publishers)? And aren’t buyers going to eventually want to buy video audiences in real-time at scale through the programmatic pipes? Shouldn’t you be getting prepared for that day?
8. I guess we’re going to have to debate Programmatic vs. Automation again. Programmatic means something that follows a program—like buying advertising using a data-driven algorithm or selling inventory based on an optimization formula designed to boost per-page yield. Automation is about using technology to simplify processes such as ordering in guaranteed campaigns or streamlining the flow of reporting data.
The two definitely go hand-in-hand, but we shouldn’t use the terms inter-changeably. We should try to recognize which technology is which, and label appropriately. Semantics has never been ad tech’s strong suit.