When launching a display campaign in March and discovering in June that half of the audience segments relied on third-party cookies that are disappearing, the realization is harsh. Teams that have not adapted their data collection and targeting processes find themselves facing degraded performance, with no immediate lever to correct it. The issue is no longer theoretical: it affects budgets, tools, and the way each campaign is constructed.
European AI Act and marketing: what it changes in campaigns
The European AI Act, formally adopted in 2024, imposes specific obligations on AI systems used in targeted advertising, content recommendation, and customer scoring. These constraints directly affect the tools that marketing teams use on a daily basis.
You may also like : Discover the benefits of lingam massage in Paris: sensuality and spiritual awakening
In practical terms, if an automated segmentation tool is used to personalize campaigns, it is now necessary to document the training datasets, manage algorithmic biases, and ensure transparency towards users. The CNIL published several guidelines in 2024 on the use of training data and the risks of re-identification in AI-based marketing tools.
For operational teams, the news from Marketingrama info helps to keep track of these regulatory developments that directly impact the configuration of tools. Documenting datasets has become an operational obligation, not a compliance bonus.
Related reading : News, society, politics: the major trends shaping today's Africa
Before connecting a scoring model to a CRM, one must verify the source of the data, test biases on sensitive segments, and prepare a clear information notice for clients. Companies that skip this step expose themselves to sanctions, but more importantly, to a loss of trust that is difficult to regain.
Contextual targeting and retail media: concrete alternatives to third-party cookies
Google has postponed and then relaunched the end of third-party cookies, but advertisers have not waited. In 2024, advanced contextual media buying has become a structured alternative to classic people-based targeting. We no longer target a user profile: we target a reading context, a moment, a device signal, or even local weather.
Players like The Trade Desk and GroupM have documented this shift. Semantic analysis of pages is gradually replacing the cookie: the algorithm reads the page content, identifies the topic, and serves a relevant ad without needing to track the user.
Retail media: transactional data takes over
The other underlying trend is the rise of retail media. Retailers with purchasing data (order histories, average baskets, visit frequencies) offer brands the opportunity to target directly within their ecosystem. Transactional data replaces behavioral data from the browser.
For a marketing team managing product campaigns, the shift is tangible. We move from web browsing-based targeting to targeting based on what consumers actually buy. Feedback on this point varies by sector, but the underlying trend is clear.
- Advanced contextual targeting relies on semantic analysis, device signals, and environmental data (weather, geolocation, time of day)
- Retail media leverages transactional data from retailers to target buyers in their purchasing journey
- Modeled data (cohorts, statistical twins) allows for broader reach without individually identifying users
Short video and AI-generated content: balancing volume and relevance
On TikTok, Instagram Reels, and YouTube Shorts, brands are producing more video content than ever. Generative AI accelerates production: scripts, subtitles, format variations. One can produce ten versions of the same video in a day.
The ground-level problem is that producing more does not mean engaging more. The algorithms of these platforms favor retention, not volume. A fifteen-second video that holds attention until the end performs better than five thirty-second videos that users scroll past after three seconds.
Editorial strategy: less content, better calibrated
Teams that achieve results in 2024 apply a rapid testing logic. They publish a short format, measure the completion rate, and iterate. AI is used to accelerate iterations, not to replace editorial thinking.
User-generated content (UGC) remains a strong lever because it brings an authenticity that brand content struggles to replicate. Combining UGC with AI-assisted editing allows for a sustained publishing rhythm while maintaining a credible tone with consumers.
Marketing personalization in 2024: how far to go without crossing the line
Personalization remains a structuring trend, but the dial has shifted. With the new recommendations from the CNIL on AI and personal data, ultra-granular personalization strategies pose a real compliance problem.
One can personalize an email based on declared purchasing behavior. One can adapt a landing page to the campaign segment. However, crossing non-consented behavioral data with AI scoring exposes to legal risks that many companies still underestimate.
- Personalization based on first-party data (forms, customer history, declared preferences): clear and compliant framework
- Personalization through algorithmic scoring: requires model documentation and explicit information to users
- Predictive personalization with third-party data: high-risk area post-AI Act, to be legally framed before deployment
Digital marketing in 2024 hinges on the ability to combine performance and compliance. Teams that invest time in documenting their tools, testing their short formats, and diversifying their data sources are the ones that maintain their results over time. The obligations of the European regulatory framework eliminate fragile targeting practices and push advertisers to structure more robust campaigns from a technical and legal standpoint.