AdMonsters new and old have assembled for another heady three and a half days of ops education and commiseration at the 41st Publisher Forum. This time, we’re in Palm Springs, CA, in palm tree-studded valley surrounded by beautiful but forbidding mountain ranges. In other words, it’s a lot like ops, minus the palm trees.
Watch this spot throughout Monday and Tuesday, March 6-7. We’ll be live blogging from the main room here at the Omni Rancho Las Palmas, bringing you a minute-by-minute report of keynotes, the State of Ad Ops, the Digital Media Leadership Awards winners’ panel, and anything else that happens in front of the full group.
Approaching 9 a.m. on Monday, AdMonsters Chairman Rob Beeler is about to take to the dais to welcome the group and introduce PubForum 41’s first keynote speaker, TEN (The Enthusiast Network) CEO Scott Dickey. Scott’s keynote, very timely considering how cross-platform distribution is shaping media consumption and the broader discourse in society at large right now, is called “Game of Platforms: How Publishers Stay Atop Content Fragmentation.”
9:03 a.m.: Rob walks out to the tune of “Danger Zone” by Kenny Loggins and promptly dispels the “rumor” that he’d gone to bed at 8:15 last night. (I can attest to that; I was in the AdMonsters secret lair/mobile office until after 11, after which he hit the hotel bar.) The goal, he contnues, is for ops to figure out how to get out of the danger zone.
9:07: Rob mentions one of the core tenets of AdMonsters: Only ops can tell ops how to do their job better. It’s not up to self-styled “industry experts.” That’s the value of PubForum, that the discussion is attendee-driven.
9:21: Hey, we have a new AdMonsters app, which is going to work for any AdMonsters events going forward! Rob has posed a live question to the group for everyone to answer in the app: Where are you currently having the most difficulty monetizing your audience? In the first two minutes of live polling, the top answer by a long shot is “mobile.”
9:23: Scott Dickey takes the dais for his keynote. TEN was a “legacy publishing company that was adrift in the transition,” he explains.
9:25: The company had great recognition, but had been bought and sold so many times. In 2007, Source Interlink and Primedia took it over with what Scott calls a good idea at the wrong time. Then the recession hit, social media exploded, YouTube came on the scene, and the company was on a “highway to bankruptcy.” In its troubles, it was owned by six banks by 2009. Those banks owned a magazine distributor that happened to own a media company. In short, it was a real challenge. So when Scott came on board… he decided it was a good idea to rebrand.
9:27: In 2013, the company got rid of the magazine distribution piece and became Source Interlink Media. But the company didn’t yet have “permission to run away from print” in spite of print’s decline. Hence the name change to TEN (The Enthusiast Network) and new focus. The business was about people who were passionate about these niche interests.
9:28: Now TEN is the top automotive and top action sports network in the space. The core of the audience is affluent males, who are influential among their peers and highly engaged with both content and advertising. There are a lot of automotive publishers out there–many of them in the room, Scott says–but TEN sees its unique angle in the influence its audience has with their peers.
9:31: There are a lot of similarities psychographically, but not demographically, among TEN’s audiences. That’s part of its play to advertisers. TEN did a study and found its audiences influenced the purchasing behavior of others and was highly engaged.
9:32: The print piece is still part of the business, and it’s a burden. One of the challenges is managing it as it declines, with the decline being essentially unstoppable as far as Scott is concerned. Their goal is to create growth in every other channel.
9:33: You need more investment from sponsors/partners in events, social media and video to grow digital channels. TEN has seen success there in terms of revenue and real audiences turning out to live events. Events may be less profitable and have a lot of overhead, but they’re an important part of the media mix.
9:36: Social is where TEN’s biggest audience is, and they’re more focused in social than they are in mobile or OTT. They have 106 million social followers in automotive–TEN’s largest audience, but lowest revenue. The question is how they can leverage that scale with advertisers.
9:38: TEN thinks Facebook Live is a promising channel; they’re trying to “crack the code” there. YouTube was an important part in TEN’s digital growth. It’s also been very profitable thanks to sponsorship and brand integrations.
9:41: TEN knew it was going to be risky to take their video and put it behind a paywall, but they had the video assets. They acquired motorsports assets from FOX along the way. They’ve launched across multiple distribution channels and connected devices. For TEN it’s a major push. Their model has been primarily ad based, and it’s important for them to drive subscriptions now.
9:43: Netflix is syndicating one of TEN’s top shows, “Roadkill.” This has led to conversations with other distribution platforms to produce content for them. TEN has a real production facility, which Scott championed. Without the ability to produce this material in-house, they would not have been able to push into branded content or to have those conversations with distribution platforms about producing new series for those platforms.
9:47: Anyone can concept a series, Scott says. Not everyone can produce it without farming out the work and losing the margins.
9:48: Source Interlink was #1 in print for automotive media in 2017. But in 2017 TEN is #1 in multiple digital channels for automotive media.
9:49: One of the challenges for TEN now is finding the right people and creating the right environment to retain those people in a disruptive, rapidly evolving environment.
9:50: Rob is on the dais right now, starting to ask questions of Scott. Rob wants to know whether events and social and these other niches have any synergy or whether they’re all separate. Scott says the company had gotten rid of some unprofitable niches before pivoting, and there’s been some churn since. But they’re “an account-based sales force that serves every tool in the chest,” he says, and the content creators are “armed to capture the content across the board in every story.”
9:52: How about measurement across all these channels, including the IRL event channel, Rob asks? Scott says the conversation has to be about delivering the best consumer experience and knowing how to distribute it.
9:54: Event people are going to do the event piece, you have content people, you have ad ops, so how do these teams work as a whole, Rob wants to know? It’s 100% about allocation of resources, Scott says, and having a core element of your business in decline gives you permission to cut. You need to think about deployment of resources. Now, TEN didn’t go into Snapchat, because the company didn’t want to allocate staff to it. So there’s a conversation about how to allocate resources over emerging channels, and diverting them from traditional channels.
9:57: “The best brands produce the best content,” Scott says. But that’s not about ads to him. On mobile, ads are to be avoided, as a user.
9:58: There are three types of people, Scott says: Those who want to believe in a new direction, those who don’t care, and the cynics. You need to get rid of those who don’t care and turn the believers into evangelists. Then you need to root out your potential believers in cynics. Some of those people can be your best evangelists if you can win them over. That helps you create the right culture and it’s good for morale. But there are lots of little dings to morale in navigating the marketplace. That’s a challenge.
10:00: How is TEN measuring across data, one attendee asks? Well, there are a lot of legacy entities, a lot of subscriptions to manage, Scott says. It’s a challenge to unify data assets from email, social, print, events, YouTube and other channels. A DMP helps. The mobile device and email together can help create universal identifiers for understanding TEN’s user base.
10:03: Rob wants to know if it makes sense to be “all in” on Facebook, whether TEN has taken any pause over the influence of Facebook and Google. Scott says there’s a real fight for market share. And TEN is also not as large as a major TV broadcast network. But content creators are connected to social and other channels. Is the experience something the user will return to?
10:09: Rob is now introducing Geoff Wolinetz, FreeWheel’s VP, Enterprise Solutions, to present FreeWheel’s sponsor session on “True Yield.” According to Geoff, that’s FreeWheel’s concept about how to balance the need to drive revenue, the buyer’s requirements, and the user experience. These elements are often at odds with each other.
10:10: FreeWheel has a lot of revenue to deliver and not a lot of inventory. Agencies are trying to prove their worth, and they’re very demanding. You have a lot of stakeholders running in different directions.
10:12: The product team gets complaints from users about too many ads. Finance has increasing revenue targets. Agencies have loads of requirements for third-party tracking to make sure they hit their target audiences. Distribution and marketing wants revenue shares.
10:14: Think less about competing demands and more about the strategy your business sets at an executive level. As it funnels down, it affects the way teams do their day-to-day work. There is a feedback loop driven by business intelligence, and it needs to feed back to the top.
10:17: Guess what? We have a user experience problem in the industry. It’s all over the trade pubs. Even Saturday Night Live is running fewer ads in linear TV.
10:18: Ad clutter, ad fatigue, latency, poor creative quality and inconsistent experiences all lead to poor ad experience. You need to get to the content quickly, as a user.
10:20: Ads per mid-roll break has gone up to 4 in 2016 from 2.7 in 2013. That’s a lot. Ads appear more than three times per stream 5% of the time and per session 11% of the time. Latency might be up to 6.5 seconds–that could be the length of an entire ad.
10:22: The opportunity is to get targeted by content, platform and audinece. Get innovative about custom sponsorships, ad placements, and pod structure. One FreeWheel client told Geoff, “If I know it’s my user’s birthday, can I serve them zero ads?” Also, understand what other teams in your business are up to and understand what agency partners want–and why. What people are asking for might not be right for what you can actually do on one particular platform or channel.
10:26: In addition to educating and empowering teams and partners, monitor data constantly to understand UX and how it’s evolving over time, and set up checks and balances in the sales workflow to flag issues before they turn into problems. Consider all sides–don’t assume the agencies or the UX or product teams you work with have different goals from you.
10:28: Rob wants to know where VAST 4.0 fits in all of this. Geoff says VAST 4.0 is interesting because it pulls out viewability and ad ID, and runs them concurrently rather than consecutively. On the client side, there’s more flexibility for measurement. But that (VPAID) introduces latency. Server-side is better for live events, Geoff says–you don’t want to be behind Twitter. Client-side works in other environments.
10:30: Rob wants to know where we are in OTT. Geoff says he was surprised mobile was 4-to-1 a concern for this room, above OTT. OTT is a more “lean-back, engage with the content” experience. Mobile is much more immediate. OTT and VOD have had problems getting traction as far as marketplace is concerned–there’s confusion over how to measure it.
10:32: Publishers and agencies know the distinction between broadcast and digital. But now there are problems with metrics when those people are thrown into the same bucket. Geoff says those issues are “starting to solve themselves.”
10:35: Nielsen is no longer the one metric that matters, Geoff says. Now you have a bunch of metrics, and you have to feed them back to the buyer based on what makes sense to them and their needs on the screen they’re working with. This especially will change as things get more addressable, more one-to-one as opposed to one-to-everyone.
12:03 p.m.: The first round of attendee breakouts are done, and we’re back in the main room for OpenX’s sponsor session, “From PMPs to RTG: The Evolution of Programmatic Direct.” OpenX GM of Programmatic Direct Paul Sternhell is presenting.
12:05: Advertisers want not just scale, but predictable scale, Paul says. So PMPs allowed buyers and sellers to operate one-to-one within a many-to-many marketplace. But the workflow is still pretty manual (one attendee chimed in that they’d describe PMPs as “programmanual”). For buyers, there’s a lack of predictability in spend, and for publishers, a lack of predictability in revenue.
12:07: There’s a discovery piece in PMPs too. An advertiser is looking for its target audience, and those audiences exist across multiple publishers. They want to know where the audience overlap is across publishers before they buy. As a publisher, you need to capitalize on the buyer’s enthusiasm. Committing to a flexibile guarantee is a new thing for publishers. That gives the buyer some wiggle room to make them happy. A sliding CPM is an even newer idea.
12:10: OpenX launched Real-Time Guaranteed, which allows for guaranteed revenue for publishers and volume guaranteed for buyer. There’s a flexibility in it that you don’t have in automated guaranteed, says Paul. And for the buyer, it’s possible to execute within the DSP.
12:12: DSPs are gradually adding functionality for RTG. OpenX is working with pubs to tighten up header integrations to better implement RTG.
1:34: We’re back in the main room after a hearty lunch, and Rob Beeler is framing the State of Ad Ops. He’s taken a “psychographic reading” from the crowd based on the responses to the questions we asked of work groups before dinner last night.
1:35. If digital media were a sitcom… One group said “Diff’rent Strokes.” Why? Because everyone does it slightly differently? “Growing Pains”–digital media is going through some. “Seinfeld”–sure, maybe a lot of it is “about nothing. “The Office”–says enough. ANd “It’s Always Sunny in Philadelphia”–it sure is chaotic.
1:37: If header bidding were a comic book character… Attendees offered Dr. Manhattan (all-powerful, but not at home anywhere), The Riddler (clear enough), The Hulk (has two sides, and just don’t make publishers angry), The Flash (“At this conference, we don’t talk about Flash,” Rob jokes–but header bidding is fast, and has good intentions but sometimes messes up).
1:39: If viewability were a pop song… We’ve got “No” by Meghan Trainor, “Work” by Rihanna, “Cry Me a River” by Justin Timberlake. These are advertisers’ responses to viewability, right? Also: “Somebody’s Watching Me” by Rockwell and “In Your Eyes” by Peter Gabriel.
1:41: If DFP were a rock star… One said Justin Bieber (can’t avoid him), Madonna (constantly reinventing), Ozzy Osbourne (so, uh, young Ozzy or old Ozzy?–answer was that Ozzy is unpredictable and never knows what to do), Keith Richards (“old, established, kind of effed up,” as that one attendee group said).
1:42: If ad ops were a character in a blockbuster movie: Jeff Goldblum in Independence Day (second fiddle to Will Smith), Q from James Bond, The Joker (several great actors who have put their own spin on it), Meliisa McCarthy in general, Rocky (the underdog who comes out on top, eventually).
1:45: The role of ops, says Rob, is fragmented in the same way media is fragmented. People’s roles are becoming siloed, and that affects what one person or team can control. If 50% is putting out fires, 50% is being strategic, and 50% is management… the math is wrong, but the situation is true.
1:47: Ops can contribute and take ownership of the way your company progresses. Change is happening throughout the industry, so how is your company evolving? How is your career developing? Where do you want your career to go?
3:25: We’re back in the main room after another envigorating round of breakout sessions. Rob is about to introduce Digital Media Leadership Awards winners Alex Jakovleski (Director of Ad Products, Twitch), Brian Brownie (Head of Business Development, Unity Technologies) and Chris Pirrone (General Manager, USA Today Sports). The DMLAs are a way of recognizing people who have helped move ad operations forward as a field, and who have helped ops people in their careers.
3:32: Chris says he’s a lawyer, and he got into tech to understand how to do due diligence for the tech company he worked for. He wanted to know whether relationships or good technology were driving their business.
3:33: Brian started out trafficking campaigns in 2000, where the industry was coming up itself.
3:34: Alex had been an anthropology and sociology major in college. No one wanted to talk to him at a career fair, he said, but a career counselor said to him, “That internet might be a thing.” He got into a search engine, then Lycos, and connections he made brought him eventually to YouTube, where he really got his ad ops chops up.
3:35: Of course, this line of conversation all falls into Rob’s line that no one goes to college to major in ad ops. It’s hard for everyone at first, Rob says, but something grabs everyone and pulls them along this path.
3:36: Alex was interested in advertising and thought he might want to be a sales person, but didn’t think he had the right temperament.
3:39: Brian explains he took what he learned from ad ops and applied it to his work at eBay, where he built out the company’s programmatic business. Chris points out that now ops is about more than yield management, it’s revenue management. USA Today and its parent company, Gannett, have given a lot more power to ad ops, which is a relatively recent thing. For a long time the sales people tried to hold back the growth of programmatic. That’s changed now. Programmatic is a major part of yield, even for a publisher with a long print history.
3:42: Alex says ops was once seen as a cost center for people to do grunt work. Now other parts of the organization ops used to face, and ops is flourishing.
3:44: Brian says there are a few different “modes” for changing your business: The abrupt mode, where you fire your sales team. Another is to teach sales how to fold in programmatic, having the conversation instead of shying away from it out of deference to how you’ve always done things. People who didn’t pass the test would be encouraged to pursue different niches.
3:45: Chris says it was important to stop calling programmatic “remnant.” Then there was the question of how to make it clear direct and programmatic were not competing with each other. Then sales people started coming to ops to try to learn something from buying trends.
3:47: Rob wants to know what people new to ops can do to help move their careers along. Brian said his mentors pushed him to speak with clients and face outward, which was not intuitive for him as a person more inclined to analyze things on his own. But that helped him advance and grow.
3:49: “What we do is figure stuff out,” says Alex. “There’s not a book for it.” Everyone is learning about how to grow as the field of ops grows.
3:57: Talking about key challenges these three are facing. Alex says getting third-party reporting on native is hard. Chris says video is a focus. It’s hard to get people to invest in something if it’s not sold. The finance people want to know a channel is not going to lose money. Programmatic can help bring in a little revenue while you’re proving video’s worth, so you’re not losing on it in the interim. Also, he says, brands just need to understand you can’t take a 30-second ad and run it in mobile, for example. Brands aren’t thinking about the experience of the medium; they just know they have assets and want to re-use them. Brian says using standards that are not necessarily the buyer’s custom standards is a challenge.
It’s 9 a.m. on Tuesday, and we’re back in the main session room, some of us a little worse for wear (by “wear,” I really mean “open bar, thanks sponsors”). Jeff Sample, who until recently was VP of Ad Operations at IPG Mediabrands and is now working with Dr. Pepper and other brands to help them take programmatic in-house, is about to take the dais to talk about consolidation in the industry and what it measnt that more brands are taking tech in-house.
9:07: Jeff Sample is asking who we have in the room–ops, product, yield, sales–there’s one sales person! Okay…
9:08: There’s a shift happening in the industry–a need for ad tech professionals to work with brands internally. That’s important for brands considering bringing progammatic in-house.
9:10: Everybody is getting smarter in digital, and it takes a lot of smarts. Brands have been bringing DMPs and first-party data in house–owning the data piece of ad tech for themselves. That’s a new thing, within the last two years or so.
9:13: It’s a natural progression to go from owning the DMP to owning the DSP. But we didn’t get here out of nowhere.
9:14: As much as it’s a buzzword, transparency is really important. Brands want it and feel like they’re not getting enough from their partners. Brands are asking whether agencies have their best interests in mind. That impacts trust. They need trust in their executing partners–that agencies will act on their behalf–and trust in themselves to execute.
9:15: Brands are in first-party data conversations. Maybe they don’t have their hads on all the tools, but they’re in the room.The more you do something, the better you get at it. Pressure from your peers can help you get better at it. Being seen publicly can help you get better at it. Brands are ramping up in all of these aspects.
9:16: The tech is getting easier for brands, too. DSPs and DMPs are being “thrown at brands” now. Those conversations with or about vendors are no longer being mediated by agencies.
9:17: So what does that mean for the business model? Publishers have historically spoken to agencies, who in turn talk to a bunch of brands. The agency is a “center of excellence” in this model. If agencies are removed, does that double or triple sales channels for publishers? It blurs the lines of communication, or makes them less linear.
9:19: Vendors also have historically spoken to agencies. Imagine how messy it’ll get if they have to speak to all the brands.
9:20: Negotiating the vendor space has been a major part of agency people’s jobs.
9:21: Companies are starting to build unique offerings. As brands get smarter, consolidation will expand further, Jeff says. He doesn’t think that’s a great thing. There’s a risk a company’s offerings will get too stretched out. Agencies really have taken a lot of the strain out of the process.
9:22: Agencies are in a tight spot, and they will have lower margins of error going forward. That’ll increase the pinch for them, and they’ll be challenged to manage it this year.
9:23: Publishers need to open up inventory via technology. Jeff is excited to talk to people about where publishers’ trouble spots are. This is something the buy side needs to know. Sales channels will be challenged and complicated for publishers, and they”ll need to manage those conversations.
9:24: It’s hard for brands to have programmtic conversations right now because there are too many vendors, many of them duplicating offerings from other vendors.
9:25: Brands will advance their tech capacity internally, and they’ll put agencies under a microscope to see how they’re performing.
9:26: Will agencies consolidate as well? Jeff thinks probably not. There’s already been consolidation, and you need talented people in agencies to do this work. Experts in how trading desks work will be in demand to work with platforms.
9:30: Gavin wants to know whether Jeff would remain on the brand side, helping them with these issues, if he had the option. Jeff says he absolutely would. There’s important work to be done there.
9:33: How about acquisitions? If you’re on the brand side, you think of DMPs and DSPs as being together. Adobe has a full tech stack. Not having to worry about cookie loss is a great boon, Jeff says. That’s very competitive, when one company has DSP and DMP capabilities.
9:35: How about the duopoly? “Facebook is Facebook,” Jeff says. When you’re in Google, though, you’re in it pretty thoroughly, because they have so many tech capabilities. Jeff doesn’t think Google is “for large brands” because it doesn’t have enough features and support for what those brands really want or need.
9:36: There are only a few independent DSPs yet. Brands probably aren’t going to bring them in-house, Jeff says. Agencies might, and Jeff is surprised it hasn’t happened yet.
9:37: It takes a lot of capacity to have conversations with all the tech players necessary to take programmatic in-house.
9:39: Gavin wants to know if groups of publishers will be at an advantage in dealing with brands. Jeff thinks it’s a great idea. It would make those conversations more efficient and worthwhile.
9:42: A lot of brands are tired of display because they’re tired of dealing with fraud or viewability, says Jeff. Now if we were to switch to a time-based model? Problem is, it requires a lot of tech. It’s pointless to make everything perfect with metrics, he says. Time can be a good supplementary metric. It would be hard to replace impressions with time.
9:45: An attendee wants to know how brands will help combat or prevent fraud, which she said often falls on publishers’ shoulders. Jeff said that as it turns out, he hears brands saying the same thing about their role–they feel like they’re bearing the brunt of the fraud problem and have an easier time just blocking. What that means is, everyone along the chain needs to be more vigilant. He says he doesn’t like to use war analogies, but it’s a battle.
9:51: Jeff compares the baseline infrastructure of the industry to the Oroville Dam blowing up.
11:02: Michael Richardson from AppNexus and Tom Levesque from Index Exchange are on the dais now, talking about server-to-server. They’re joking that they’re frenemies and need space. In reality, Index and AppNexus have agreed to support S2S mutually.
11:04: Everyone’s heard competition between exchanges is good for publishers. It’s done away with the waterfall. And it’s still new, even though we’ve been heavily in it for 18 months or so. The consequences, though, will go on. It’s time to move beyond experimentation and toward optimization.
11:05: Header bidding is a lot of work for the browser. That’s the downside. We’re getting a lot of best practices, but there’s a lot that can go wrong in a header implementation. Can we take the basic premise and apply it to other environments, aside from the browser?
11:08: Ideally header bidding would be transparent and configurable, plus fast and supported in all environments. This is a “maturation thing,” says Michael–we need to think about what works in the long term.
11:11: S2S isn’t new. This is how DSPs and SSPs communicate. The SSP knows how to put code on the page. What’s new is that the lifting moves away from the page. S2S is a pure tech layer. A lot of companies are providing both demand and a tech layer, and they need to stop worrying about who’s going to outbid whom if they’re going to exist in this tech layer together.
11:15: About the Index/AppNexus partnership: Michael says people are going to be seeing a lot more of this. These two companies are trying to signal to the industry that they’re willing to mutually support S2S. Tom says they’re also trying to get publishers thinking about their criteria for a S2S partner.
11:18: You want a hybrid header/S2S approach, to be able to test partners where it makes sense–should the partner be on the page or in a S2S layer? Tom says it ought to be configured with a switch, not a major implementation effort to move bidder around.
11:22: Rob says he thought the benefit of moving everything to S2S is that you have all your reporting in one place. Michael says there are pluses and minuses. Being on the page gives you control. As the space evolves, there may be more incentives (beyond light load) in moving toward S2S. The hybrid approach will dominate for some time, Tom says. Your best performing partners can live on the page, while the “long tail” can go to the server side.
11:25: You have your tech layer and your demand layer–make the distinction between them. You need to think about what your partners bring to the table to avoid coming upon a new kind of partner overload. If you’re on the fence about where to put your bidders, lean toward whatever creates a better user experience.
11:32: In partner evaluation, ask: Of your partners, who can provide the pipes? Who’s integrated with them? If someone is telling you to prioritize one demand over another, that damages the entire point of header bidding. The partner should bring no bias.