David Joosten shares a privacy-first roadmap for marketers—activating first-party data, resolving identities, and using AI to drive growth in a cookieless world.
As the era of third-party cookies fades and privacy regulations tighten, marketers are facing a critical inflection point: how to build data systems that are compliant and capable of driving sustained growth. In this interview, David Joosten, Co-founder and President of GrowthLoop and co-author of First-Party Data Activation, shares a pragmatic roadmap for designing a privacy-first marketing data stack, regardless of company size.
From identity resolution and data minimization to the decisive role of generative AI in audience generation and journey orchestration, Joosten explains why first-party data is the foundation for future-ready marketing. He also highlights how smaller organizations can actually benefit from building privacy-by-design systems from the ground up. Whether you’re rethinking your tech stack or planning for an AI-enabled marketing future, this conversation offers grounded insight and actionable advice for navigating today’s fast-evolving landscape.
Full interview;
What are the critical components of a privacy-first marketing data stack, and how can even small or mid-size companies begin building toward that vision?
The foundation of a privacy-first marketing stack rests on three critical components, which we detail extensively in the book:
- A consent management platform (CMP) that goes beyond simple cookie banners and instead provides granular control so that customers can consent to specific uses, like analytics versus marketing emails.
- A marketing data lakehouse that combines the flexibility of data lakes with the performance of data warehouses. This combination creates your single source of truth to break down the data silos that plague most organizations. It also serves your analytics and AI.
- An activation solution, such as a composable customer data platform (CDP) that connects directly to your marketing data lakehouse rather than creating another copy, dramatically reduces costs and complexity.
For smaller companies looking to enable self-serve capabilities for marketers, I always recommend starting with the “crawl, walk, run” approach from Chapter 3. This allows marketers to avoid the need for outside consultants or additional hires.
The key insight is that you don’t need complete or perfect data to start. Focus on data minimization from day one—only collect what you truly need, ensure it’s encrypted in transit and at rest, and build your identity resolution on deterministic matching using email addresses or customer IDs. Small companies have an advantage here because they can build privacy-by-design from the ground up, rather than retrofit legacy systems.
Also Read: Are B2B Marketers Finally Moving Beyond Last-Click?
What role does identity resolution play in successful first-party data activation, and how should marketers think about choosing or building an identity graph?
First-party identity resolution matches disparate data points about a customer into a unified profile, which is critical to marketing results and regulatory compliance. Without it, companies miss the opportunity to unify data points about a customer into a holistic customer journey that helps them personalize their outreach. This hurts relevance and campaign performance.
More importantly, poor identity resolution is more likely to lead a company into regulatory troubles. In today’s privacy-centric environment, false positive matches are far more dangerous than false negatives. A false positive means you’re potentially matching data between different people—that’s a GDPR violation waiting to happen.
The biggest mistake I see is marketers jumping straight to complex probabilistic matching when they should start with simpler, rules-based approaches. Begin with your most trusted data source (usually your purchase or transaction data), and carefully layer on additional data sources in your marketing data lakehouse. Test extensively as you make changes.
Consider your organization’s needs and strengths when deciding whether to build or buy an identity resolution solution. If you have a strong analytics team and care about transparency, build deterministic matching in your marketing data lakehouse—it’s often more reliable and easier to audit. A CDP with built-in probabilistic matching may be a good fit if you’re a product-led company with a high volume of anonymous web traffic.
How do you see generative AI and large language models integrating into the modern marketing stack beyond content creation, especially in campaign planning and customer journey orchestration?
This is where the future gets really exciting. In Chapter 9, we explore “compound marketing,” the exponential growth that comes from AI continuously improving your campaigns at a pace humans can’t match.
Beyond content creation, we’re seeing three transformative applications:
- AI-powered audience generation where models can analyze all your customer data, channel performance, business context, and marketing goals to suggest audiences you’d never think to create. We’re already seeing this with platforms like GrowthLoop, where marketers can simply say. “Help me reduce 90-day churn,” and the Audiences Agent builds a variety of segments to target.
- Predictive journey orchestration, where AI doesn’t just follow pre-built customer journeys but dynamically creates personalized paths for each customer. Instead of “If email doesn’t work, try SMS,” we’re moving toward “For this customer, based on their behavior patterns, the optimal next touchpoint is an in-app message in 3 days.”
- Real-time campaign optimization, where AI agents can interpret campaign performance results to make targeting and creative adjustments faster than any human team. This allows you to iterate and scale from dozens of simultaneous campaigns to thousands of personalized interactions.
The most clever use of AI is testing messaging and content on artificially generated audiences. Rather than running expensive focus groups, marketing teams can generate detailed insights on how their customer personas will feel or react to their content.
But here’s the critical insight: AI without quality first-party data is expensive guesswork. The organizations that win will be those that have built strong data foundations first. That’s why we emphasize the composable architecture throughout the book—your AI capabilities need access to all your customer data, not just what lives in one platform.
In the book, I discuss the current shift from “marketer as creator” to “marketer as advisor,” where you set strategy and guardrails while AI handles execution. This shift is not about replacing human creativity but amplifying it at an unprecedented scale.
Also Read: Why Brands Must Build Parasocial Bonds, Not Just Content
Your new book, First-Party Data Activation, is a timely guide for marketers navigating a privacy-first future. What inspired you to write it now, and who did you have in mind as the ideal reader?
As martech professionals, my co-authors and I kept hearing the same fundamental questions from marketing and technology leaders: “What does a modern marketing tech stack even look like anymore?, “How should we think about AI’s effect?”, and most importantly, “How do we prepare ourselves—and our teams—for what’s coming next?” The data, strategies, and technologies we relied on just a few years ago have quickly become outdated. The old world of third-party data, once the backbone of digital marketing, is quickly becoming unreliable. Privacy regulations are tightening (as they should), customer expectations are shifting dramatically, and AI is no longer a futuristic concept—it’s here, actively reshaping how we work. The marketing technology landscape has exploded to over 14,000 solutions, yet most teams are drowning in complexity rather than driving results.
Again and again, we found that the answer wasn’t in some cutting-edge tool or secret growth hack. It was something much more fundamental: first-party data. Not just owning it but truly understanding how to use it effectively. It’s about moving beyond the basics: rethinking data strategies, embracing privacy-first approaches, and learning how to activate data to drive real, measurable effects to compound growth.
We started searching for a guide—something practical, strategic, and grounded in real-world experience. But we couldn’t find the resources we needed, so we chose to write one ourselves.
We designed for marketing leaders, data-driven strategists, and marketing technology professionals who are looking for clear, actionable frameworks in a time of constant disruption. We released it now so marketing and technology leaders can read it over the summer before their upcoming annual planning cycles.
This isn’t about keeping up with trends. It’s about building marketing data platforms that compound value over time rather than requiring constant manual effort to maintain performance.
One of the book’s core themes is helping marketers modernize their data stack. What are companies’ most common mistakes when activating first-party data, and how can they avoid them?
The biggest mistake is “analysis paralysis”—waiting for perfect data before taking action. Companies spend months debating data quality while their competitors are already running personalized campaigns with incomplete but actionable data. The truth is that you need to build momentum for these first-party data programs by generating marketing wins and a return on your investment along the way.
The second major mistake is platform proliferation without integration. Companies add a CDP, then a personalization engine, then an attribution tool, creating what we call “Frankenstein’s data stack”—multiple systems that can’t talk to each other, each requiring its own data copy, making targeting less accurate.
Here’s a specific example: a major university we worked with used Tealium for identity resolution, but when they needed transparency for regulatory compliance, they realized they had no visibility into how matching decisions were made. They were paying for a black box when they needed an explainable system.
To avoid these pitfalls, we recommend keeping your data stack simple and extensible:
- By simple, I mean investing in a single source of truth for your first-party data in a common data platform like Google Cloud BigQuery, Snowflake, etc. When new systems come online, everything should flow back to that single source of truth. You should also activate all campaigns from that source.
- By extensible, I mean that you can easily add and remove technologies or cloud capabilities because your data is not siloed in any of your downstream tools.
Also Read: What Does Modern Customer Experience Look Like in 2025?
You often talk about turning marketers from “project managers” into “agile experimenters.” What does that shift realistically look like, and how can marketing leaders encourage it?
Great question, and this transformation is at the heart of everything we advocate for in the book. Right now, most marketers at larger organizations spend 80% of their time coordinating between systems—filing tickets with IT, waiting for audience exports, manually checking that suppression lists were applied correctly, etc. They’ve become project managers of their campaigns rather than growth drivers.
Here’s what the transformation looks like in practice: Instead of filing a ticket and waiting two weeks for an analyst to create an audience of “customers who abandoned carts in the last 7 days,” a marketer can define that segment themselves and have it activated across email, paid social, and push notifications on the same day.
To encourage this shift, marketing leaders need to do three things well. First, invest in the right architecture to give marketers the ability to self-serve with the correct guardrails (e.g., the composable CDP approach we detail in Chapter 4).
Second, establish experimentation as the default—every audience should have a control group, and every campaign should measure incremental lift, not just vanity metrics.
Third, and crucial, change how you measure success. Instead of asking, “Did we hit our open rate targets?” ask, “Which test has had the biggest impact on customer lifetime value?” For example, we’ve seen teams discover that their highest-value segments prefer SMS over email or that personalized subject lines work for acquisition but not retention.
The practical reality is that this requires new skills. Marketers need basic data literacy—not coding, but understanding concepts like statistical significance and incrementality testing. They also need to be comfortable thinking across channels to create a holistic customer journey. But the payoff is enormous: instead of optimizing individual campaigns, you’re optimizing the entire customer experience.
For example, we worked with Indeed’s marketing team, and they went from launching 2-3 audiences per quarter to 8x that volume after implementing self-serve capabilities. But the bigger change was qualitative—marketers started thinking in terms of hypotheses to test rather than campaigns to execute.
As marketing becomes the enterprise’s most “technical” department, what new skill sets or mindsets do you think will define the next generation of marketing leaders?
This is such a crucial question because we’re witnessing a fundamental shift in marketing leadership. The next generation of CMOs won’t just be creative visionaries—they’ll be data-driven growth architects who can navigate technology and regulatory decisions that directly impact customer experience and business outcomes.
There’s no doubt that a deeper understanding of data, including an understanding of analytics and insights, will be essential for the future marketing leader. Despite the ability of AI to help you interpret results, it still will fall to marketing leaders to think critically about the patterns of results to set guidelines and direct AI-driven marketing systems effectively.
Brand, design, and storytelling are as relevant as ever. AI can augment and scale your team’s creative thinking, but cannot help you find or define your brand’s soul and why it exists in the world. I believe recognizing “world-class” level versus merely “good” design, storytelling, and brand will continue to be essential.
Finally, future marketing leaders will tremendously benefit from a growth mindset focused on constant experimentation, both for campaigns and the underlying technologies their teams use. They will spearhead wave after wave of transformations driven by consumer demands, regulations, and technology changes. As a result, future marketing leaders must lean into constant learning and test the limits of the possible.
Practically, this means marketing leaders need to be comfortable conducting data reviews with the CFO, technology architecture discussions with the CTO, and compliance meetings with the legal team. They are becoming the bridge between customer insights and enterprise capabilities, which makes marketing one of the most strategically important functions in the organization.
