In one of our previous editions, we discussed the different data privacy regulations around the globe. Despite some differences, they have the common purpose of protecting user privacy. Another radical change that has affected the advertisement world is the 3rd party cookies policy, which will be regulated according to the privacy laws, as well as the latest iOS implementations that curtails advertising data.
The new regulations provoked concerns in the advertising industry. Marketers are no longer able to collect user data in the same way as previously. The privacy data regulations demonstrated how vulnerable the marketing industry actually is. For one, the billion-dollar company Facebook in just a few months suffered a double-digit drop in advertising revenue, meanwhile, their targeting is continually becoming less accurate. Moreover, most MarTech companies still rely on cookies in some form. Google announced back in January 2020 that it would remove 3rd party cookies from Google Chrome in 2023 (the date has been postponed in the meantime). So what will happen in the industry in 2022?
Cookies are a relic from the desktop world. They are a “simple technology” used to collect extensive data about users’ online behavior, which advertisers can use to learn more about their customers and develop targeted offers for them.
However, customers are increasingly concerned about who and what, in what way, and what kind of information is collected and sold. The main problem with cookies is not personalization: customers have little objection to needs-based advertising and product offers. Their objection lies with privacy and a lack of control over their data.
Adello has been developing targeting technology that does not rely on cookies since inception in 2008. The idea that mobile marketing deserves a better solution than the prevailing “desktop goes mobile” approach, which relies only on cookies and clicks, stems from founder Mark E. Forster’s background in app development and his interest in analytics and data. Not relying on cookies proved to be the right move in the long term, considering that cookieless approaches are now the focus of all major MarTech companies.
Fortunately, authorities around the world respect user privacy; Apple, one of the industry leaders, has moved to handle advertising data more strictly in the newer iOS versions. For cookies, this means the end in the long run.
What are the expected changes in a cookieless world, and how will 3rd party cookies policies look like?
The new regulation requires asking users about their consent before storing or accessing information. The principle of user consent is freedom, specification of information, and it has to be based on an explicit affirmative action. EU Data protection authorities are aware that 3rd party cookies restriction can complicate the work of businesses, which used this technology before. Therefore, authorities released recommendations on the alternative methods to get user consent.
Many companies have already started reworking their websites. According to the report, cookies are down by 22% across news websites. However, the majority of the websites are still using 3rd party cookies and do not meet the GDPR requirements yet. Advertisers working with non-compliant publishers put themselves at risk.
The end of cookies does not necessarily mean that advertisers have to move away from personalization entirely. The challenge is to find an approach that is both efficient and ethical.
The good news is that there are several alternatives. Adello, for example, has never used 3rd party cookies but developed AdCTRL™ starting in 2011. The name derives from the objective at the time: End users should have control over whether and what kind of advertising is desired. Originally, an app was developed for iOS and Android where in cooperation with all publishers the consent and rejection to advertising should be shared. However, since there was no cooperation with the publishers (respecting that users may not be shown any ads at all proved to be a critical point), the technology pivoted towards sustainable data collection for precise targeting.
Here are the main targeting alternatives:
For probabilistic methods, the systems detect different sorts of data points in order to identify the personality as best as possible. This is usually achieved by blending a combination of observed behavior with “soft identifiers”.This includes device type, publishers visited, on-site behavior, and the IP address. This method aims to provide tracking and attribution.
Deterministic data is difficult to collect. The challenge of the probabilistic method now is to identify a unique user, which in turn is not privacy-compliant. It’s a fine line that companies like Drawbridge in the U.S. and, more recently, 1+X in Switzerland are walking.
Such solutions are mostly “non-cookie based” and include a combination of methods. Nevertheless, one thing is clear: Probabilistic methods are technically challenging, and the results are not always correct. Many customers have rejected probabilistic methods in recent years.
In the partnership with Unified ID 2.0, Trade Desk released their own solution for the post-cookies era. Since popular web browsers such as Safari, Firefox, and Google Chrome rejected the cookies, Iteration 2.0 will replace the use of 3rd party cookies with hashed and encrypted email addresses.
The developers claim this ID will remain open and, moreover, will implement significant upgrades to consumer privacy and transparency. This framework can be used by all companies, not just those working with The Trade Desk.
However, this method mirrors the functionality of cookies too. That is the reason Google criticizes Unified ID 2.0, claiming it might be a not sustainable long-term investment.
The Advertising ID Consortium is an open and independent group, which is operated by representatives from well-known AdTech companies, such as Index Exchange, LiveRamp, The Trade Desk, dataxu, and others.
The Advertising ID Consortium provides privacy-conscious and people-based interoperability for the advertising ecosystem. The method they suggested is cookies ID utilization via AppNexus’ domain as well as a people-based identifier supplied by LiveRamp’s Identity Link. The idea is to hash login-email data. Also, apply the same hashing method whenever a user receives an advertising email and clicks a link. That hashed information will create an ID that will be captured in local storage and will be turned into a 1st party cookie.
Apple has a range of solutions and one of them is SKAdNetwork (or SKAN). SKAN allows accurate attribution of app installs. This framework includes three main entities:
Publishing app, the ad network which signs the ad, and the app advertisement. The important thing to understand is that the entire process is unrelated to Apple’s device identifier for advertisers (IDFA).
After iOS 15 was released, new versions of SKAdNetwork appeared (now in 3.0). Developers and marketing measurement platforms, such as Appsflyer, Adjust, Kochava, Branch, and others, are using SKAdNetwork, to install validation postbacks, and perform their campaigns on their own.
There are several disadvantages of SKAN. Firstly, it only works for in-app, and secondly, it is designed for iOS. Moreover, the data presented is limited to the 100 campaigns for an app or a network. However, Adello uses SKAdNetwork for in-app solutions and sees this framework as reliable and future-proof.
Adello AdCTRL™ relied from the very beginning on user consent and observed human behavior in order to maintain privacy. It complies with all applicable data protection laws, including the European Data Protection Regulation (GDPR), the California Consumer Privacy Act (CCPA), and the Brazilian General Data Protection Law (LGPD).
Currently, the methods of data processing within AdCTRL™ fall into three main dimensions:
1. Content (publishers). Traditional and proven media planning relied on the selection of relevant environments. Adello extends this with the latest technology, continuously searching and classifying publisher data (mobile sites, apps), so that the system understands the content’s environment, keywords, and ratings. Based on this metadata, it can be checked whether an app is trustworthy or not. Does the app have only 2-star ratings? Is the app is Indonesian, but the ad space is offered in German? The system continuously maintains a content quality index based on different, statistical models. The detailed understanding of content, quality, and environment allows better targeting, is robust and absolutely privacy compliant.
2. Context (contextual, temporal, and spatial data). It’s all about when and where an ad is placed. For example, the time when the ad is posted may very well be relevant for an ad for coffee. Furthermore, the place where the ad is placed contains additional data such as the current weather or the type of location, which can tell something about the behavior of the device in question (Adello has drawn polygons of all major airports, train stations, and most retailers and includes such data). Is the location a coffee shop, an office building, or a residential area? That could play a significant role. This method relies on identifying legitimately available 3rd party data (e.g. weather providers, static data e.g. from OpenData/Statista/Federal Statistical Office) and combining appropriate data with its own data (e.g. polygons) to create an understanding of locations. Thus, the algorithm knows in which situation an advertisement will be served.
3. Behavior: The last dimension is the manner, of human (inter)action. This goes far beyond the click. It starts with technographic information (which device is used with which firmware and browser), through connection speed, response, and dwells times, to interact with content and ads (i.e. swipes and clicks). This provides a good understanding of interests and relevance, i.e. currently observed human behavior, and is used in combination with the two methods described above in real-time. This, of course, requires user consent.
AdCTRL™ was developed for processing large amounts of data in real-time, and patented accordingly. The latest generation of AdCTRL™ provides a robust and future-proof targeting framework - while fully preserving privacy.
Adello Instant Classification AIC is a new, simplified version of the more elaborate classification in AdCTRL™. The main difference is that AIC leverages the power of in-stream data processing to instantly classify users into just three categories (age, gender, interests) without having to draw on all of AdCTRL™’s available historical data. The before and after are largely masked out. This reduction results in a coarser but faster identification with similar accuracy - it is generally superior to current probabilistic methods.
There are currently several alternatives, and there will appear more as soon as companies become privacy-conscious and demand cookieless solutions. The elimination of 3rd party cookies is not the end, but on the contrary, the new beginning of a more ethical, sustainable, and privacy-friendly approach to marketing. Choosing the right tools and adapting them will save you time and increase your competitiveness.
A data strategy that focuses on all customer touchpoints and leverages current technologies is a good first step and important for sustainable digital marketing. In the future, we might well see how giving up cookies might have been a blessing in disguise: the inflection point which allowed the industry to finally achieve fast and qualitative growth.