Pour one out for programmatic marketers. They’re struggling to get by, and not just because of the ongoing, recurring depreciation of third-party cookies.
There are stock quality issues raised by recent controversies surrounding sites designed for advertising, Forbes And Colossus. There’s a mess of brewing measurement as third-party cookies eventually disappear. And of course there’s AI, which may well be one of programmatic marketers’ biggest potential allies in tackling some of the other challenges – although AI poses its own problems.
At the Digiday Programmatic Marketing Summit last week in Palm Springs, Calif., agency executives participated in closed-door public discussions, during which they were granted anonymity for their franchise. And they’ve been very candid about the challenges they face. Here is a sample of what was said.
AI problems
“I think if you asked 90% of agencies what AI is, they couldn’t give you a clear answer. Each business has its own meaning.
“If you just say AI is going to come to you, the inputs will be bad, the outputs will be even crappier, and fraud will take over.”
“We have a full-time job running our campaigns, running our programs, and partnering with clients. Then, also solving the AI problem, just because we are dynamic nerds and able to talk about it, is not a good enough reason for us to tackle it.
“It’s a question of R&D. Agencies are mainly funded for performance and purchasing. So you have a fundamental problem of financing, of training these models.
“We’re renewing contracts with some customers, and some are reporting, ‘Hey, whatever you’re using AI for needs to be explicitly stated. And now there are caveats like, “Hey, creative, you can’t use AI here.” » Because there is a certain level of fear. Inclusivity is not built into AI.
“We don’t put any financial data or anything that compromises the customer’s identity (into AI tools).”
“There’s always that human element – it’s not like I’m just going to take the result and go, ‘Oh, I don’t need to think about that at all.’ For me it’s a time saver, but it doesn’t replace me still having to do my job to ensure accuracy and quality and all that kind of stuff.
“The data privacy angle is why we don’t have (an AI chat instance) open at the moment, as we probably will in a few months. This is because we don’t use ChatGPT due to the data they have. Our intellectual property, we are audited. ChatGPT, we can use it for personal purposes, but it is no longer allowed in our organization. We leverage Gemini (from Google) because they have the rights and privacy, so we are able to maintain them as part of our relationship with Google Drive.
“A conservative company that’s being audited is very hesitant to put anything, like cost information, into an AI unless they know it’s a closed bubble where competitors or bad actors cannot gain access.”
AI Effectiveness
“Every time my team cuts and pastes from one place to another, it’s a last-mile solution that AI and automation need to help fill so my team spends less time trying to get the job done and more time trying. to improve the work. »
“I’m happy to take charge of AI development and seed ideas for my team, as the efficiencies allow us to achieve more competitive pricing.”
“A great example of an AI project I worked on was we just created a chat instance of all aspects (for campaign creative sizes). Because every time you launch a campaign, the creative team asks themselves: “What size banners do you need?” And they never change.
“There is software called CorralData. I can ask: “Which store is the best performing among these 200 franchise stores?” And it will aggregate all the data (and) in about five seconds it will start spitting out answers. This saves me a lot of time.
“We’re working with a casino and the creative team created (an AI chat instance) that helps them name casino promotions based on whatever they’ve already named a promotion, words they don’t will never use. And it doesn’t replace creative hours, but it at least helps them refine their project, instead of coming to clients and saying, “Oh, we already did that four years ago.” »
“We work with Seedtag and we use their generative AI creations to create additional creatives for some of our brands that don’t have as many. And of the campaign we tested it on, it was the only one we had that generated impact, consideration and notoriety.
“You can also use (AI) in identity matching. If you have a Ramp ID, UID, Google ID, and integration (conversion API) on Meta all trying to claim the same individual, you can ask (the AI) to s Practice this to create associations to help you refine whether you’re reaching someone in a fragmented location or taking them to another area, to help with your outreach approach.
Cookie exhaustion
“Obviously you’re still working to meet the deadline for cookies to be obsolete, but I think people are past that deadline, to be honest.”
“Everyone is talking about it, but no one is thinking about the real implications of what’s going to happen on the campaign because it’s been so long (since companies started preparing for the deprecation of third-party cookies).”
Measurement issues
“In terms of targeting, we are in a good position. Many agencies and brands are well placed there. Where we struggle is measurement and attribution.
“A lot of our dollars go to the metas and Googles of the world because they can say, ‘I put in $1 million and got $6 million out.’ Man, that’s a great story that my finance team understands. But that ties our hands because a lot of our dollars are going there because of the gap on the measurement side of being able to prove from a clear, perfect line.
“If I have to tell my CMO, ‘I hired the best agency, the best team, but I can’t tell you what I was telling you (in the measurement reports),’ that’s a huge problem that our industry must find a way to solve and work together.
“The customer needs to be aware of the changes coming and how they also need to evolve and invest.”
“When we started, I felt like digital was a performance guide; it wasn’t a real one-on-one. And that turned into a last click (attribution model), but there are so many different factors that trigger that outcome.
“We always knew that there was data on people and media, and then on the elements that emerged from your campaign. And then (agencies) started buying the data companies, and now they’re all in one place. But that doesn’t mean the same thing. That doesn’t mean you can put it up. It’s like trying to put Magna-Tiles and Legos together. We are now stuck in a situation where we have our Magna-Tiles and our Legos and we do this every day (hand gestures as if trying to combine opposing forces with each hand). And we know it won’t work, but they’re paying us to do it.
Inventory issues
“There is a lack of trust. We are contacting partners who lied to us a few years ago (about how they managed the programmatic supply chain). You’re going to talk to people and some may directly lie to you or not know the answer themselves. So it’s a meeting we have, but it’s a painful exercise.
“There is so much CTV inventory being resold now. You may be buying from Magnite on a PMP, and I don’t want to say anything bad about Magnite, but don’t put CTV trash in the deals feed.
“(THE Controversy over Google video partners) was a huge scandal. Buyers who bought YouTube were actually buying Google Video Partners, and the service works on God knows where. They say it’s not MFA, but we pulled a URL and there were some interesting sites on there.
“(Supply path optimization) should be an ongoing conversation.”
“In one of the cases, specifically Forbes, we are talking to some brands, as well as GARM and TAG, that in fact (Forbes) knew about it. And so were they classified as regular monetization? So, if it was under their domain header, if it was under Forbes.com, which it was, is it really MFA? »
“(SPO) is constantly evolving, and we are always evolving. This is not something that could be considered settled. It must be reviewed every year, for each account, for each brand.
“We’re just playing whack-a-mole with all this fraud and cheating.”
“I don’t like feeling stupid, like you did something wrong because we’re not watching it closely enough. That’s crazy.
“We pay (verification companies) to protect our media. And now we need to find another solution to protect the media again. Where to draw the line? And what are we sold to?
“It’s our job as a community to push back (verification companies) and say, ‘Okay, so you’ve created an AI solution that you think is going to help us solve this problem; how do you teach it? How do you update it?’
Executive 1: “It’s incredibly manual (you have to go through campaign reports to check which URLs contain ads). But we must do it.
Exec 2: “This is why we need AI. »
