DAC Blog Authors Attribution reimagined: Probabilistic methods in privacy-first marketing
Filter By
Healthcare Analytics and Marketing Science Services Content Strategy Customer Relationship Management Design and Creative Services Digital Media Local Listings Management News Paid Media SEO Strategic Insights Web Development COVID-19 Series See all our authors
Digital moves fast.
Subscribe to our monthly newsletter to get ahead of the curve with new articles, videos, white papers, events, and more. Unsubscribe anytime. For more information, see our Privacy Policy.
Attribution reimagined: Probabilistic methods in privacy-first marketing

Attribution reimagined: Probabilistic methods in privacy-first marketing

Monday, April 15, 2024

As digital marketing undergoes a transformative shift away from its traditional reliance on deterministic data and third-party cookies, new opportunities to innovate are being revealed—particularly in the realm of attribution. The move towards a privacy-first approach, driven by tighter regulations and evolving consumer expectations, challenges us to rethink how we attribute success to our marketing efforts.

This period of change isn’t just about adapting to new privacy norms; it represents a pivotal opportunity to innovate in how we measure, understand, and optimise marketing impact. By exploring privacy-focused, probabilistic methods and leveraging advanced technologies like Google’s Privacy Sandbox, we’re not only navigating a new era of digital marketing but also pioneering more sophisticated, ethical, and effective ways to achieve attribution.

Current challenges in digital marketing

The digital marketing world is rapidly changing due to strict privacy laws like GDPR in Europe and CCPA in the US, which limit how consumer data can be used. The demise of third-party cookies, once a digital advertising staple, has shaken up the scene, with major browsers cutting them out. This shift forces marketers to look for new, privacy-friendly ways to engage their audience. The industry is now moving towards using first-party data and probabilistic approaches that respect user consent and anonymity, a big leap from previous cookie-reliant tactics. This evolution demands marketers to be more adaptable and educated, as they work with new tools and strive to deliver the personalized experiences consumers expect, all within a more regulated framework.

The rise of media mix modelling (MMM)

Media mix modelling (MMM), sometimes also referred to as marketing mix modelling, is a powerful analytical tool that quantifies the impact of various marketing strategies on sales. By analyzing historical data, MMM reveals how different advertising channels, like TV, digital, and print, contribute to sales, helping allocate marketing budgets more effectively. It identifies high-ROI channels, optimising spending for better results.

MMM is particularly valuable today as it respects user privacy, operating with aggregated data rather than personal details, making it ideal in privacy-conscious settings. It provides a comprehensive view, accounting for variables like market trends and competitive actions, enabling strategic decisions beyond immediate returns.

At its core, MMM uses data to inform smarter marketing choices, reflecting the shift towards strategies that are both effective and privacy-compliant. It empowers businesses to navigate marketing complexities, ensuring investments are impactful and aligned with privacy norms.

The practical applications of MMM

Applying MMM in your business involves a structured approach to measure the impact of marketing channels on sales and other key business outcomes.

  1. Define goals: Begin by setting clear objectives for using MMM, such as optimising marketing budgets, enhancing ROI, or analyzing the effects of marketing on sales.
  2. Collect and prepare data: Compile data on sales, marketing efforts (both digital and traditional), economic conditions, and other relevant factors, covering at least two years for comprehensive historical insights.
  3. Prepare your data: Clean and format your data, fixing any issues with missing data or outliers, to make it suitable for analysis.
  4. Build and test your model: Employ statistical methods like regression analysis to create a model that elucidates the relationship between marketing activities and business results. This step may require experts in statistics and possibly coding.
  5. Analyse and optimise: With your model in place, examine the outcomes to determine the efficiency of different marketing channels, using these insights to reallocate budgets towards the most effective channels.
  6. Update regularly: Market conditions and channel effectiveness fluctuate. Updating your MMM periodically with fresh data keeps your marketing strategy in tune with current trends.

MMM’s applicability varies by company size. Startups might use MMM for crafting growth strategies using industry data and metamodelling. Small to medium businesses (SMBs) can leverage MMM to stretch marketing dollars further for growth, while large enterprises can refine their marketing strategies across multiple channels for optimal results.

Implementing MMM requires a blend of marketing acumen and statistical analysis skills. For organizations lacking these capabilities internally, there are MMM software and platforms available that automate much of the data handling, modelling, and analysis process, making it more accessible.

Google’s Privacy Sandbox initiative

Google’s Privacy Sandbox is a comprehensive response to growing privacy concerns, aiming to balance the need for personalization in advertising with the increasing demand for privacy on the web. Central to this initiative is the development of new technologies designed to phase out third-party cookies in Chrome by providing alternative methods for ad targeting and measurement that do not compromise user privacy.

The initiative introduces several key technologies:

  • Topics API: This technology aims to deliver relevant ads to users by categorizing their browsing history into broad topics of interest, without the need for third-party cookies or identifiers that can track individuals across different websites.
  • FLEDGE (First Locally-Executed Decision over Groups Experiment), now referred to as Protected Audience: This facilitates remarketing to users by allowing advertisers to target ads based on users’ previous website visits, while keeping the data on the user’s browser to prevent personal information from being shared across sites.
  • Attribution Reporting: Provides data linking ad interactions (clicks or views) with conversions (such as purchases), without compromising user privacy.
  • Private Aggregation and Shared Storage: These technologies allow for the generation of aggregated data reports and provide a mechanism for websites to store ad-related data locally in Chrome, offering privacy-preserving access to this data.

Additionally, the initiative includes Fenced Frames, which enable the secure embedding of content on a page without the risk of cross-site data sharing, addressing security and privacy concerns associated with traditional iframes.

The Privacy Sandbox’s development has been marked by Google’s collaboration with the industry and regulatory bodies, such as the UK’s Competition and Markets Authority (CMA) and the Information Commissioner’s Office (ICO), to refine and test these technologies. This collaborative approach aims to ensure that the solutions developed meet both privacy and functional requirements for the web ecosystem.

Despite these technical advancements, the transition to a model without third-party cookies has elicited mixed reactions from the industry. Some view these changes as a necessary evolution towards a more privacy-respecting web. Others, however, express concerns over Google’s growing control over web standards and the potential implications for competition in digital advertising.

As the Privacy Sandbox technologies continue to evolve, it remains a focal point of discussion around the future of privacy, advertising, and the open web, reflecting a broader industry shift towards enhancing user privacy while maintaining the ad-supported internet ecosystem.

Leveraging Google’s Privacy Sandbox for future-proof advertising

For marketers, the practical application of Google’s Privacy Sandbox technologies means preparing for a future where personalization and privacy coexist, so it’s important to proactively engage with the new ecosystem rather than passively waiting for changes to occur.

  1. Get familiar with key technologies: Focus on the Topics API, Trust Token API, and Privacy Budget API. These tools facilitate interest-based advertising, combat fraud, and limit personal data access without breaching user privacy.
  2. Strategize and allocate budget: View third-party cookie deprecation as an opportunity. Align strategies with privacy advancements and dedicate resources to new technology trials.
  3. Use privacy-preserving APIs: Explore APIs like Topics for targeted advertising and Attribution Reporting for ad effectiveness, emphasizing privacy.
  4. Test and collaborate: Engage in testing and provide feedback. Collaboration shapes these technologies to better suit marketing needs. Google stresses the importance of collaboration for refining these tools, so early engagement can influence their development to better meet your needs.
  5. Innovate for a cookieless future: Prioritize first-party data and machine learning to adapt to privacy-first advertising trends.
  6. Stay informed and be proactive: Regularly update your knowledge of Privacy Sandbox initiatives through Google and industry news. Staying at the forefront of developments allows for agile adaptation to new advertising norms.

Adapting to a post-cookie world

The shift away from cookies to a new paradigm demands a radical rethink in data collection and use. The focus must be on transparently gathering first-party data, leveraging probabilistic targeting, and optimising cross-channel strategies through MMM.

In this evolving digital landscape, the imperative for marketers to innovate and adapt has never been more pressing. After all, embracing a privacy-centric, probabilistic approach is not merely a technical challenge; it’s a transformative journey. Marketers are tasked with staying ahead of the curve, continuously updating their knowledge on the latest tools and regulations, while championing ethical marketing. Leading the change doesn’t just mean keeping up with the times—it means redefining the standards.

Fortunately, you don’t have to get there alone. Talk to the experts from DAC’s data analytics center of excellence, Proove Intelligence, and find out how we can prepare you to succeed in digital marketing’s next chapter.


Subscribe to our monthly newsletter to get ahead of the curve.
Get exclusive access to new articles, videos, white papers, events, and more. Unsubscribe anytime. For more information, see our Privacy Policy .