We have one for programmatic marketers. They're struggling with this problem not only because of repeated deprecations of third-party cookies.
There are inventory quality issues caused by the recent controversy surrounding Forbes and Colossus, sites made for advertising. The eventual retirement of third-party cookies is causing measurement disruption. And of course, there's AI. In fact, AI may be programmatic marketers’ greatest potential ally in addressing other challenges. However, AI also has its own predicaments.
At last week's Digiday Programmatic Marketing Summit in Palm Springs, Calif., agency executives participated in a private town hall discussion and spoke on condition of anonymity to discuss their candid opinions. And they were very candid about the challenges they had to face. Below is a sample of what was said.
AI problems
“If you ask 90% of agencies what AI is, I don't think you'll get a clear answer. Each company has a different meaning.”
“If we just sit there and say AI is going to happen to us, the inputs will be worse, the outputs will be worse, and fraud will be rampant.”
“We run campaigns, run programs, and partner with clients full-time. And solving AI problems requires more than just being dynamic nerds who can talk about AI. That’s not a good enough reason to work on AI.”
“This is an R&D issue. Agencies are primarily funded for performance and procurement. So you have the fundamental problem of funding to actually train these models. .”
“We're renewing contracts with some of our clients, and some of them have warned us that 'anything that uses AI needs to be clearly stated.' And now… , with a warning that says, “Creatives, you can't use AI here.” People are scared because of the level. Inclusivity is not built into AI. ”
“We do not publish financial data or anything that compromises the identity of our clients.” [into the AI tools]”
“There's always the human element. I'm not just going to take the output and think, 'Oh, I don't have to think about this at all.' For me, it's a time saver, but it's not something I It doesn't replace it. I still have to do my job to ensure accuracy, quality and all that.”
“From a data privacy perspective, we [an AI chat instance] We're opening now, but probably within a few months. That's because ChatGPT isn't using the available data. Our intellectual property is audited. ChatGPT is available for personal use, but is no longer available for organization-wide use.we take advantage of [Google’s] Because Gemini owns the rights and privacy, we can keep it as part of our relationship with Google Drive. ”
“Conservative companies that are audited are very hesitant to incorporate things like cost information into their AI unless they know it's a closed bubble that's not available to competitors or nefarious actors.”
AI efficiency
“Whenever my team is copying and pasting from one place to another, AI and automation are the last-mile solution to fill in the gaps. [that] My team now spends less time getting work done and more time making it better. ”
“We are happy to take on AI development and ideas for our team, as the efficiencies allow us to achieve more competitive pricing.”
“A great example of an AI project I worked on was creating a chat instance for all aspects. [for campaign creative sizes]. Because every time we launch a campaign, our creative team asks, “What size banner do I need?” And they never change. ”
“There's a software called CorralData that says, 'Out of these 200 merchants, which one is the best performing?'and aggregate all the data [and] Start spitting out answers within 5 seconds. Doing so will save you a lot of time. ”
“We are working with casinos and the creative team [an AI chat instance] This will help you name your casino promotions based on all the words you've ever named your promotions with, words you'll never use. It won't replace your creative time, but it will at least help you refine it instead of going to your client and claiming, “Oh, he already did that four years ago.'' ”
“We've been working with Seedtag and using their generated AI creative to build additional creative for some brands that don't have much of their own creative. Out of all the campaigns we had, this was the only one that was able to get more lift and consideration and awareness.”
“You can also use [AI] In adjusting identity. Ramp ID, UID, Google Identifier, and [Conversion API] Meta-integration, which attempts to all claim the same individual, [the AI] Train it to create associations and adjust whether you're contacting someone in a fragmented location or taking them to another area to help you approach your reach. ”
cookie burnout syndrome
“Obviously, we're still working toward the deadline for cookies to be phased out, but I think people have gotten over it, to be honest.”
“Everyone's talking about it, but it's been so long that no one is thinking about the actual campaign implications of what's going to happen.” [since companies originally started preparing for third-party cookie deprecation]”
measurement problems
“We're in a good position when it comes to targeting. A lot of agencies and brands are well-positioned. What we struggle with is measurement and attribution.”
“A lot of our money is going to the metas and Googles of the world because they can say, 'We put in $1 million and got back $6 million.' Well, my finance team understands that. But it ties our hands because a lot of our money is spent there because of the gaps on the measurement side that can prove to be off the clean, perfect line.”
“If I were to say to a CMO, 'I hired the best agency, the best team, and now I can't say the things I used to tell you,' [in measurement reports]This is a big problem that our industry needs to find a way to solve and work together to tackle. ”
“Clients need to be aware of the changes that are coming and how they must evolve and invest.”
“When we first started, we felt that digital was the guide to performance. It wasn't really one-on-one. And then it turned into last click. [attribution model]But there are so many factors that cause that outcome. ”
“We've always known that there's data about people and data from media and data that comes from campaigns. Then [the agencies] We started acquiring data companies and now we have it all in one place. However, they do not mean the same thing. That doesn't mean you can do it by combining them. It's like trying to combine Magna Tiles and Lego. We're now in a situation where we have Magna Tiles and Lego and we do things like this every day. [gestures with hands as if trying to combine opposing forces with each hand]. And we know it doesn't work, but they pay us to do it. ”
inventory issues
“Trust is lacking. Reach out to the partner you lied to years ago. [about how they were managing the programmatic supply chain]. You're going to talk to people, but some may lie to your face or you may not know the answer yourself. So this is a meeting that we do, but it's a painful exercise. ”
“We have a very high inventory of resold CTVs right now. You may be buying from Magnite in your PMP. I don't want to say anything bad about Magnite, but I don't want to put garbage CTVs in my bid stream. Please.”
“[The Google Video Partners controversy] It was a big scandal. The buyers who were buying YouTube were actually buying Google Video Partners, and God knows where that would run. It says it's not MFA, but I pulled the URL and found some interesting sites. ”
“[Supply path optimization] It requires an ongoing conversation. ”
“One of them, especially Forbes, we're really working with some brands, GARM, TAG. [Forbes] I knew about it. So were they classified as regular monetization? So if it was under the domain header, if it was under Forbes.com, is that actually MFA?”
“[SPO] is always evolving, and so are we. It's not something that can be considered resolved. Every account, every brand, needs to be reviewed annually. ”
“We’re just playing whack-a-mole with all the fraud and fraud.”
“I don't like feeling like an idiot, like you did something wrong because we weren't watching enough. It's whack-a-mole.”
“We pay [the verification firms] To protect our media. And now we have to come up with another solution to protect media again. Where do we draw the line? And what are we being sold to? ”
“As a community, we have a responsibility to resist this.” [the verification firms] “Okay, so you've created an AI solution that says it helps address this problem. How do you teach it? How do you update it?”
Performer 1: “It's incredibly manual.” [having to crawl through campaign reports to check URLs that carried ads]. But we have to. ”
Executive 2: “That’s why we need AI.”
