There’s been tons of chatter in the industry at large about signal loss, but what does signal loss mean to the ecosystem, specifically publishers?
For better or worse, Publishers are always left bearing the brunt of industry shifts. With addressability shifting into something different, many wonder how we will continue to target and reach our audiences.
The cookie phase-out process is here, and word on the street is that Chrome will be obfuscating the IP address (the cookie of CTV) real soon.
During a panel titled “The Future of Addressability: The Portfolio View,” Anthony Katsur, CEO of IAB Tech Lab, sat down with Shobha Doshi, SVP of Programmatic Strategy & Operations at Raptive, Ryan McConville EVP of Ad Platforms & Operations at NBCUniversal, and Mike Nuzzo, SVP of Hearst Data Solutions at Hearst Magazines. Each panelist outlined how they are approaching addressability today.
One thing we all know to be true is that there is no one solution; publishers should instead consider the “patchwork quilt” of solutions that are at their fingertips. With a catalog of over 5,000 publishers, Doshi highlighted that Raptive is currently in the test iterative stage. She encourages publishers to continue testing and exploring to see what works and what doesn’t.
At NBCU, they are rebuilding their signaling around first-party identity. With Peacock and their other digital endpoints totaling around 300, they had to find a way to coordinate those varying identities across everything. Having recently launched the NBC Unified identity platform, their strategy is to elevate it to live primarily on first-party identity signals.
When asked how Hearst thinks about the future of addressability, Nuzzo kept it simple. “We’re doing a lot of testing, learning, and just making sure we’re following the law,” he shared. From his perspective, the onus has been on publishers to solve for innovation. Innovation and identity are increasingly challenging because, again, there is no one solution.
Future-Proofing Alongside Hefty Privacy Constraints
Privacy regulations are evolving in Europe, and while a handful of states with state-led privacy regulations are already in place, three more states will be enforcing privacy sanctions next year: Montana, Oregon, and Texas. India also just passed a privacy law.
What alternative solutions are publishers seeking to maintain an addressable ecosystem while complying with privacy regulations?
“I think it’s hard, and that’s why I said the legal piece earlier,” Nuzzo explained. “Consent management platforms are in a good place right now, we need them contextually.” At Hearst, the focus is on understanding common taxonomies and how they apply that to their audiences algorithmically. Gen AI is also something they are pushing towards.
For McConvile and NBCU, the product team is essential to maneuvering the privacy landmine. Media and entertainment companies have privacy product managers who enforce all privacy regulations. Rather than creating nuances for each state’s privacy laws, they look to the states with the most conservative ones and use those as a baseline.
Privacy and big tech are the two big bad wolves of the industry, but which one is scarier?
It’s no surprise that Nuzzo from Hearst says it’s the legal side. No publisher wants to come out of pocket and pay off the government for not having the right privacy policy. McConnvile went with big tech, considering the different platform policies that publishers must abide by.
“Operating system policies like Apple’s override our terms and conditions, so if you sign up for Peacock on an Apple device and agree to NBCU’s terms and then opt out of Apple, it overrides our terms and conditions completely. I don’t think users actually understand that they’re making that choice,” McConvile explained.
The IP Address Is the Cookie of CTV
While CTV doesn’t operate on cookies, it does use IP addresses for audience targeting. This makes IP addresses the cookies of CTV. One or both will put a velvet rope around the web and obfuscate the IP address.
What will happen next as the IP address becomes effectively deprecated by big tech?
Doshi, SVP of Programmatic Strategy & Operations at Raptive, thinks the IP address signal loss would reduce graph strength for everyone. “It makes regulatory compliance hard, especially if there are different state-by-state regulations that may conflict,” she explained. “There are no good solutions right away, and solutions will differ per environment since regulations differ per environment. We may see some advantages on desktop, but the long road ahead is to figure out how to make it work.”
Nuzzo had an opposite opinion, “I think the IP address for us opens up an opportunity that we haven’t explored as publishers,” he said. Also, stating that there are actually a lot of good use cases for the deprecation of the IP address that we haven’t thought of. “I hope the industry allows us to explore that before they cut off at the knees,” he said.
The IP address powers a lot of how performance TV works since viewers don’t click on their televisions to buy something. The IP address connects the devices in your household. “You have a smart TV that lives on an IP address, and you have a mobile phone that lives on that IP address,” McConville explained. “So if you see an ad on Peacock on your smart TV, you can buy the sneakers from that ad on your mobile phone.”
The industry will have many holes to fill if the IP address is deprecated because, aside from targeting, the IP address is crucial for cross-device measurement. There is also some interesting work on an Internet service provider level to future-proof the ability to tie IP signals to a deterministic household in a privacy-compliant way.
From a fraud vector perspective, the IP address is used for many fraud detection and data security issues as you start to proxy through a single IP address or VPN on the web. This opens up many fraud and data security issues for advertising and health, tech, finance, and national defense implications.
In short, IP addresses have many uses in the advertising ecosystem beyond targeting. Many publishers will have to think through and plan for the many implications.
Is First-Party Data the New Oil in Our Industry?
When it comes to first-party data, we are somewhat in a world of the haves and have-nots. The walled gardens have had identity for many years, but everyone outside of them had an alternative architecture through cookies they could function on.
Publishers and brands have to create a value exchange to get the data. There are tons of data-rich companies, like Amazon, for example. You are unable to use any Amazon entity without logging in. Publishers need to make sure that their consumer product teams are creating a system for authenticating users and communicating the value exchange to garner more logins.
Now is the time, more than ever, for companies that have not traditionally collected first-party identity signals to figure out smart ways of doing it. “On the cohorting side over at NBCUniversal, they have done some really interesting tests with seed data using AI and content.
“We fed our content into an AI engine, scanned all the contextual metadata for all the content, and then created lookalike models using that more granular data set. By using this AI deep contextual metadata, which is also our first-party data, we found that the segments performed much better,” said McConville.
At Hearst, there are tiers of data assets at their disposal, so it doesn’t have to be explicit logins. They collect 4 trillion data points on their users monthly, which is precious data. Recently, they conducted a study and saw a 140% increase in click-through rate when they applied both contextual and behavioral into one segment for the advertiser.
“We’ve lived on this behavioral journey for so long, where the mentality was just follow the consumer around, and they’ll buy my shoes,” Nuzzo said. “This may be true, but when you serve 5000 impressions to them versus if you serve them the right content, you’ll have to do far less of that, and the interaction rate will improve.”
Where Do We Go From Here?
As an industry, we have a nasty habit of waiting until the last minute before we react to changes. We have built a solid muscle on the third-party cookie, so is there a sense of urgency in Q1 of 2024? Probably not.
Patrick McCann, SVP of Research at Raptive, also led a main-stage discussion highlighting how more publishers need to start testing the Privacy Sandbox. Raptive hopes that we will discuss how the buy side is ingesting and understanding some of that data next year.
Clean rooms were also discussed, and McConville predicts that we will see some real commercial action coming out of Clean Room integrations because there was a long time when they weren’t being utilized. With Amazon and Google PAIR, we see real commercial examples of bringing first-party data service into these data clouds.
Nuzzo thinks we will talk about gen AI in the spring of 2024 and finally have some real learnings from it. We will finally have some metrics to see what works, and what doesn’t work.