In early March Google has put out an announcement that set the marketing industry and its technology providers under renewed pressure to change the way the online world works. In summary, it says that “Third-party cookies are dying, and first-party datasets are becoming more valuable”. This announcement will come with significant changes and challenges, there will be winners, and there will be losers. Our belief is; if you have your first-party data under control, you will be one of the winners in this development.
Is Google’s announcement the end of targeting in digital marketing?
March 3, 2021, Google release a supposedly ground-breaking innovation announcement, which caused waves in the advertising industry: To maintain the data protection requirements of the users and regulators, as of early 2022, Google products will no longer be supporting third-party cookies.
But is this announcement new?
Not really. Regulators have had the debate around data protection on the web since the turn of the century. Due to the most recent debacle involving Facebook and Cambridge Analytica, most users are aware that many providers can build profiles of them, by tracking their online movements and behaviour data. In early 2019, Apple made the first move to cut out third-party cookies on its Safari browser, using its “Intelligent Tracking Prevention” mechanism. Mozilla followed shortly afterwards with updates to its Firefox browser. And in April of the same year, Google announced the death of all third-party cookies. In 2021 they are merely repeating what was already said back then.
Nevertheless, with Google’s Chrome browser commanding almost 70% of the browser market share, the current announcement carries more weight and ultimately calls on all players in the market to act. This debate is how user-centred communication will be carried out in the future – without leaving out the users and their consent.
In early 2019, Apple made the first move to cut out third-party cookies
on its Safari browser, using its “Intelligent Tracking Prevention” mechanism.
Mozilla followed shortly afterwards with updates to its Firefox browser.
And in April of the same year, Google announced the death of all third-party cookies.
What will happen in 2022?
Strictly speaking, it is not 100% clear how exactly what the alternatives will look like, or when they will take effect. But two points are clear:
- In the media arena going forwards, users will be hidden in groups of “similar” users, in so-called cohorts. This means: if an advertiser works with the hypothesis that there could be a disproportionately large number of potential customers in a particular cohort, they will use this cohort to address all users as a group. However, this also means that the advertiser would not know if they reached the individual potential customers or how often the individual user would have seen their ads or message.
- There are opportunities to use first-party data to improve the targeting. How exactly this will be mapped, remains open. Google only says this much: “First-party relationships are vital”. The Chrome development team is currently working on the necessary interfaces to be able to transfer first-party data.
Graphic 1. Googles FLoC (Federated Learning of Cohorts)
How do you use first-party data now?
Currently, the interface and the processes for handling data have not yet been fully developed or finalised, and the discussions are currently going on about the exact “how” this will work. The “what”, however, is clear: Remarketing as a use case must continue to be possible. There is too much value in this mechanism for the entire industry.
In the future, it will be important to think carefully about which user’s, advertisers would consider as valuable on their website, and which are not. If a user is still on their website, we can identify and address them as an advertiser, at the same time, assigning them properties. Up until now, the misdirected investment in a wrong decision might still be manageable – now thousands of (paid) contacts may be related to the question of whether a user of a cohort belongs to a relevant target group or not.
In the future, it will be important to think carefully about which user’s,
advertisers would consider as valuable on their website, and which are not.
As a result, the model used to calculate the target group membership of a user will have to improve. The aim is to find out how high the conversion probability of each user is. This has to be done with statistical methods, which is both sufficient complexity and AI-supported. A simple “What you see is what you clicked” will no longer suffice!
The second crucial point will be to make this decision exactly in the timeframe in which you can still identify a user. With the end of third-party cookies, this must happen on your first-party pages. “Real-time” is therefore becoming a decisive feature of future solutions.
At Smart Digital we are prepared for these developments in a few areas:
First-party data collection
With our combined technology and tools from other providers, we always collect first-party data in the centre – only with the consent of the user. We know from numerous tests that first-party data provide the greatest value contribution when it comes to predict the interests of the individual user correctly. In many places, this already overcompensates for the result of third-party data that has been bought in.
Improvement of user identification
Instead of the sole focus on (first-party) cookie IDs, we have already included other (first- and third-party) identifiers in our technology. We recognize users based on their logins, CRM IDs, external cookie IDs or other identifiers and merge previously separate journeys. From the combined profile, we calculate the “overall result” for the individual user. This means that third-party cookies are already unnecessary in many places (e.g. with cross-site user recognition).
Segment assignments are often hand-made. Every Amazon user knows this: once you click on the wrong product, it follows you on every further page. At this point, it is worth investing in providing marketers with an AI-based solution. With statistical models, segment affiliations can be calculated incorruptibly and free of views or opinions. We also consider that every user may think a little differently. Currently, efficiency gains from this approach are already up to 30%. The effect is likely to be further intensified by the concept of cohorts.
We already offer solutions that recalculate the segmentation of a user with each new interaction of a user’s action. Suppose Google introduces its Chrome updates in line with their announcement. In that case, we will be able to use the newly provided APIs (see also Google’s “FLEDGE”) to transfer updated segments for remarketing use cases for each interaction. Making it possible to outperform current solutions, depending on the final solution design of Chrome.
First-party data is the most important component of the data-driven toolbox
Developments in the future of digital marketing are in full swing at this point. It affects all browsers, regulatory interventions (ePrivacy) and providers of alternative technologies (e.g. ID alliances, data partnerships, etc.). One thing remains clear, the way we work with user-approved, first-party data will be the key to communicate to our audiences, in a targeted and data protection-compliant manner. Without this, advertisers could slip back to the early online advertising years of wastage investments with little return or knowledge of what worked and what did not.
To prevent this, companies need to start cleaning up their database as soon as possible. The changes in the coming years will be very complex and different in how we use the first-party data available to us. The sooner we adjust to this mindset and put in place a well-structured first-party database, the greater the competitive advantage will be in a world without third-party data.
Foto: cgterminal| Shutterstock