The end of the third-party cookie is nigh. The world's three biggest browsers—Google's Chrome, Apple's Safari and Mozilla's Firefox—have now taken measures to ensure it will be so. So how will marketers target their online ads going forward?
There's a few options available—online IDs, panel-based attribution, browser segmented-audiences, contextual targeting—but they are imperfect and come with their own challenges. Fortunately, Google has given advertisers a two-year deadline to get their ducks in a row, and most advertisers have been (or should have been) planning for this, given its inevitability amid a broader privacy shift.
Furthermore, being a mobile-first region gives Asia-Pacific an advantage. Cookies have limitations in mobile web and aren't used at all in-app, so other targeting methods have become common. Indeed, industry experts discussed earlier this week how if any region is “ready” for a shift away from third-party cookies, APAC is the closest.
So how will the industry respond to the imminent death of the cookie, and what solutions are likely to gain traction? Campaign asked a handful of experts for their thoughts.
1. Would a panel-based solution work? What are the challenges?
Kenny Griffiths, managing director APAC, MightyHive
The challenges of a panel-based solution are common to the methodology, and well recognised—extrapolating results based on a limited data set results in probabilistic outcomes, which are sure to be less precise than deterministic approaches. One of the biggest issues we face as an industry is how addicted advertisers have become to the deterministic possibilities digital advertising has afforded thus far, particularly 3P cookies. But as we all know, that precision has come at the cost of user privacy.
Probabilistic methodologies are still very valuable—traditional channels such as TV and radio have succeeded by way of probabilistic panel-based measurement since their inception, spearheaded by the likes of Nielsen. The reality is that there is a trade-off to be found between insight and privacy. The biggest challenge is in education. Advertisers must understand the varying methodologies used to produce the vast array of metrics afforded to them, both online and offline, and factor the advantages and disadvantages of the methodology into their decision making.
Niraj Nagpal, director of business development APAC, Iponweb
A panel based solution would face four fundamental challenges to its effectiveness:
1. Attribution would not be functional at a viable level as a panel essentially is counter to the value proposition of digital’s measurability on a more granular, 1:1 level.
2. Limited scale: would a panel be able to accurately extrapolate large, multicultural and geographically dispersed audiences such as those found in APAC?
3. User incentivisation: how to get users to install something—most likely a browser plugin—that would track their online behaviours more closely than third party trackers, especially as privacy is now a consideration factor.
4. Data efficacy: do people interact online the same way as if they are not being tracked?
Jonathan Beh, CEO, China, Cadreon
Yes, and we are seeing great efficiency from this. Of course there are challenges (ie scaling, as good panel-based solutions are hard to find and usually lack the data). However through this approach we experienced that our precision media targeting has been taken to the next level, from precision to prediction! One of the data sources from our proprietary people-based data stack (AMP – Audience Measurement Platform) enables us to take on the panel-based data. This is where we work with our industry panel-based solution partner’s data (we sometimes call this attitudinal data) to create predictive models to help our clients to uncover high value audiences (HVA) through a machine learning clustering technique. We have applied similar practices with our partner’s platforms, such as JD (one of China’s ecommerce giants) as well as UnionPay and have found efficiency in recognising audiences at scale through lookalike approaches. Of course, rules of engagement apply when working with these partners to ensure that all of our practices are compliant with the China’ data protection and privacy regulation.
This article is filed under...
2. Developing a common identity has been touted as a solution, how will this work?
Jason Barnes, chief revenue officer APAC, Pubmatic
This puts identity firmly front-and-centre on the adtech agenda. The move will impact almost all use cases for the third-party cookie, from targeting to measurement and attribution. In particular, I believe this change will see the digital advertising ecosystem step up its efforts to find a consumer-friendly audience addressability solution backed by first-party user login data or some type of deterministic ID.
There are a lot of premium publishers out there with user login data and ID solutions who can help publishers activate that data for marketers. We think there is an opportunity to move the open internet to leverage identities backed by user data while staying privacy compliant and this will help level the playing field with the walled gardens, to some extent. A big upside to this move is it should lead to a lot more collaboration across all players in the ecosystem over the next two years.
A challenge specific to this market is the lack of large-scale cross-country identity solutions, such as LiveRamp in the US. However tech partners like PubMatic who have strong relationships with publishers can work to aid adoption of ID solutions—we see our recently launched Identity Hub as a means of facilitating this.
It will require both collaboration from vendors within the industry, with standards to be developed and driven with industry trade bodies. The Trade Desk’s Unified ID Solution saw over 30 vendors come together and adopt the Unified ID to improve cookie synching/match rates for their own benefit, as well as their peers and competitors, and although that solution will be affected by Chrome’s intended changes, it is an example of the industry working together to create a solution to benefit the wider ecosystem. Over the coming months the industry will see vendors and industry bodies coming together to develop initiatives to solve for identity management, targeting and measurement in a privacy-safe way. And Google has already announced the Chrome Privacy Sandbox.
As this is a cross-industry challenge, I believe that most ecosystem participants would be happy to join and support any industry initiatives for solving the cookie problem. Nonetheless we will continue our own research and work on solutions that provide the tools and services needed by our clients.
From China’s perspective, we are more concerned about the “hard to get” DeviceID, mainly due to iOS hiding such information and in the near future Android devices will do so too.
Open identity link tech is already available in China for us to help our clients with people-based media and marketing solutions. It’s a very important tech because this enables us to have a holistic overview of the consumer's journey follow by a closed-loop analysis for all of our media campaign activation. All of this is done through the tracking of DeviceID.
We are still speculating whether the China personal data privacy protection law will consider DeviceID as PII; if they do, it will be sensitive and inappropriate to work on DeviceID moving forward. I guess when one door closes, another opens!
We are however working with our partners and industry players to jointly promote OAID (Open Anonymous Device Identifier), this will safeguard our media offerings for our clients which for the time being heavily rely on IDFA/IMEI. This is still early stage and not yet fully adopted by major publishers. Some good progress, as MMA China (trade bodies) already recognised and authorised such approach.
3. Do you expect advertisers in APAC will look to find workaround solutions to the cookie ban, or will they be satisfied with contextual targeting?
Indeed, concerns around ad targeting and user privacy contributed to Google’s announcement. There is every reason to believe that targeting will still be possible, as will attribution, but the mechanisms will need to radically change. Deterministic targeting at scale will likely become obsolete and platforms and advertisers may turn to federated learning, contextual targeting, and other techniques to drive performance through programmatic platforms. Another suggested approach would be for the browser itself to segment audiences based on their browsing behaviour, and once there are a sufficient number of other browsers in this interest group an advertiser could target them.
What about frequency management? In October 2019 Google introduced frequency management across bid requests without a third-party cookie associated with them. Instead, Google employs machine learning to analyse behaviour from across their ad inventory and then estimate with a high degree of confidence the number of impressions an individual had been exposed to.
Lastly, publishers with first-party audience relationships are poised to fill in audience targeting gaps left by third-party data. For example, this would include a publisher with a paywall that requires a user to login to read content. Publications are likely to sell more curated inventory packages, much of which will be available programmatically via private marketplaces (PMPs) and programmatic direct/guaranteed deals.
We should see a real push to developing machine-based alternatives that leverage privacy-compliant signals in the bidstream, such as context, IAB category, geo, and device type, to predict user cohorts or audience propensities, behaviors, and conversion probabilities. In addition, we should also see an increase in the usage of first-party publisher data to target their audiences on their own properties. Once these tools mature in sophistication and accuracy, they will allow audience-based targeting that delivers strong results for advertisers and preserves CPMs for publishers, while still maintaining consumer privacy as dictated by the browsers. Google has already indicated intentions for creating its own sandbox of tools to help with this, but there need to be alternatives in the marketplace that create choice and distinct advantage for publishers, brands, and agencies.
Rather than rely solely on contextual advertising, the industry needs to accelerate movement to user identities backed by user login data and reduce reliance on third-party cookies. Solutions like Identity Hub can help publishers move to that world sooner, and we will push as a platform to enable publishers to leverage alternative IDs and bring scale for marketers. However a better future requires collaboration from both sides of the industry. We need to bring everyone together and stop looking for band-aid solutions for the open web—instead develop together the path forward that puts the open web on par (and above with content) with walled garden’s audience addressability capabilities.
China is mobile and app driven, hence contextual advertising application will be limited, simply because apps are less open versus a website. Perhaps we need to look at contextual targeting a bit differently to the west (ie text recognition based on the web page’s content). In China we are working with a tech partner that uses AI for text, NLP/speech regognition and image (still and moving) recognition. For example in live TV, the tech can recognise a scene from the motion picture (i.e. someone holding up an iPhone) or voice (someone mentioned iPhone), and a relevant ad can appear alongside it.
China’s BAT (JB) publishers are very much dominating the scene here and continue to build their walled garden; ie. Alibaba recently introduce their Alibaba Cloud data platform, which essentially looks into bridging offline (usually via QRCode scan) with online data. There are identify linking solutions in place within this platform, purely to match the offline customers online. Tencent last year introduced similar setup via. Tencent Data Cloud.
As China is such an ecommerce-driven market and an abundance of clients already have a direct to consumer strategy in place with Alibaba Tmall for example, working and partnering with relevant publishers makes sense, because this enables them to have access to greater data as well as audience visibility. However the rules of engagement apply, which usually means no data ever comes out from Alibaba.