The Synthetic Shift: How Digital Twins are Revolutionizing Media Planning

The marketing industry is currently undergoing a fundamental transformation, moving away from reactive analysis toward predictive simulation. At the heart of this shift is the rise of the "synthetic consumer"—AI-generated audience personas and digital populations that allow CMOs, media planners, and creative strategists to test marketing hypotheses in virtual environments before spending a single dollar of real-world budget.

This development, while nascent, is gaining significant traction among major agencies and global brands. From Goodway Group’s recent partnership with Arima to the strategic initiatives at WPP, Dentsu, and Code & Theory, the mandate is clear: the future of media buying lies in the ability to simulate human behavior at scale.

The Core Concept: What is a Synthetic Consumer?

Synthetic data, often referred to as "digital twinning" or "audience cloning," involves the use of generative AI to create high-fidelity simulations of specific consumer cohorts. These models are constructed using vast, anonymized datasets—ranging from national census figures and labor statistics to Nielsen survey results and, crucially, a brand’s own first-party data.

Unlike traditional consumer personas, which are static descriptions based on historical averages, synthetic populations are dynamic. They can be queried. If a marketer wants to know how a specific demographic in Germany might react to a 15% increase in share-of-voice during a holiday season, they no longer need to rely solely on past campaign data or expensive, time-consuming market research. They can simulate that scenario against a synthetic model to observe predicted outcomes.

Chronology of Adoption: From Niche Tool to Industry Standard

While the current wave of interest feels sudden, the foundational work has been ongoing for several years.

• Early Adoption (2020–2022): Agencies like Code & Theory and The Brandtech Group’s Jellyfish began integrating AI agents and synthetic personas into the creative process, primarily to guide the development of assets from the initial pitch to production.
• Expansion Phase (2023): Dentsu partnered with startup Evidenza to begin embedding synthetic data directly into its media planning processes. During this time, the focus shifted from creative guidance to programmatic media strategy.
• The Scaling Pivot (2024–Present): 2024 has seen a surge in investment. WPP began testing synthetic audiences based on charity donation data to inform programmatic buys. Most recently, Goodway Group entered into a strategic agreement with Arima to model the responses of U.S., British, and German populations to media activations.

The Strategic Imperative: Why Now?

The move toward synthetic modeling is driven by three primary challenges: the erosion of third-party cookies, the rising cost of market research, and the need for greater agility in fragmented media landscapes.

Patching the "Data Gaps"

Paul Frampton-Calero, CEO of Goodway Group, views synthetic data as a structural necessity. "If you’re trying to understand a particular product line or a particular region, the data is usually quite limited," he explains. Synthetic models act as a form of filler for these informational notches, allowing agencies to conduct incrementality tests in simulated environments before verifying them with small-scale, real-world media activations. This minimizes waste and maximizes the accuracy of media plans.

Accelerating Speed-to-Market

The efficiency gains are perhaps the most compelling argument for CMOs. Joy Doyin Ogunneye, global innovation and brand communication lead for the Kenvue skincare brand Aveeno Body, notes that her team previously spent up to 30 days developing a single strategy brief. By utilizing Vurvey Labs’ synthetic research, that process has been compressed to just three days. This speed allowed the brand to accelerate the development of its successful "Sensitive Skin Scents" launch earlier this year.

Qualitative Depth

Crucially, these tools are evolving beyond mere quantitative averages. Chad Reynolds, CEO of Vurvey Labs, emphasizes that while Large Language Models (LLMs) are excellent at providing the "top of the bell curve" or the most common answer, their tools are designed to surface the nuances that LLMs often miss. By incorporating qualitative video surveys from real consumers into the synthetic model, they capture the "why" behind the "what," providing a more holistic view of consumer sentiment.

Official Responses and Industry Perspectives

The industry reaction remains a blend of unbridled optimism and professional caution. While the potential for speed and cost-saving is undeniable, the risks associated with "garbage-in, garbage-out" data models are a constant topic of discussion.

Julie Clark, SVP of diversified markets, media and entertainment at TransUnion, highlights the critical need for foundational integrity. "You have to start with the right foundation," she warns. Clark points out that marketers must be wary of personas built on unreliable data.

Furthermore, as agencies move toward autonomous buying agents, the industry is grappling with new operational anxieties. Clark reports hearing from clients concerned about the potential for automated systems to "blow through" budgets if left unchecked in high-concentration environments like live streaming. "Agencies are certainly grappling with what that looks like. We’re in a test and learn phase. It must have happened if people are worried about it," she notes.

Implications for the Future of Marketing

The implications of this shift extend far beyond simple media buying.

From Media Buying to Holistic Strategy

Frampton-Calero’s ambition for Goodway Group is to push these models into the realm of product strategy and pricing. By simulating how populations respond to price sensitivity or distribution changes, agencies can transform themselves from service providers into true business consultants.

The "Omnipresent" Customer

As Stef Hoffman, head of enterprise transformation strategy at Code & Theory, points out, these tools make the customer "omnipresent." When a brand can query a digital twin at any stage of the creative or strategic process, the distance between the consumer and the boardroom is effectively eliminated.

The Persistence of Human Judgment

Despite the technological leap, the consensus among experts is that synthetic data should inform, not replace, human judgment. Hoffman warns that teams must maintain a "professional distrust" of the findings. The tool provides the intelligence, but the strategy must still be crafted by people who understand the broader market context—a nuance that even the most sophisticated AI cannot fully replicate.

Market Context: A Changing Landscape

The adoption of synthetic data occurs against a backdrop of a shifting economic environment. Fresh figures from WARC indicate that global ad spend is projected to grow by 11.9% to $1.34 trillion this year, though that growth is expected to moderate to 8.4% in the coming year.

Suzy Young, head of WARC Media Data, characterizes this as an "increasingly uneven economy," where AI capital spending is propping up growth even as households face sustained cost-of-living pressures. In such an environment, the ability to optimize spend with synthetic precision isn’t just an advantage—it is a competitive necessity.

Conclusion

The era of the "synthetic consumer" is no longer a futuristic concept; it is a tactical reality. As agencies and brands continue to refine their models, moving from public-domain data to proprietary, first-party-led personas, the accuracy of these simulations will only improve.

While concerns regarding liability, model bias, and the potential for over-automation persist, the trajectory is clear. The marketing industry is moving toward a future where the digital twin is an essential colleague—always available, infinitely scalable, and capable of answering the most difficult questions before a campaign is ever launched. For those who can master the balance between synthetic insight and human strategy, the next decade of media buying promises a level of efficiency and predictive power that was previously unimaginable.