07.03.2025

The Future of Marketing Operations: Evolving AI Agents & Products Part 2: AI-Hacked Marketing at the Forefront of Evolution

2025 is being referred to as ‘The First Year of AI Agents,’ reflecting the growing attention on their impact in business. From the early stages of AI adoption, Dentsu Digital has championed the use of AI in marketing, helping client companies drive bothbusiness growth and operational efficiency. In the second installment of the “ AI-Hacked Marketing at the Forefront of Evolution” series, three specialists from Dentsu Digital’s AI division will provide frontline insights into how AI agents—including solutions like “∞AI”—are driving transformation and evolution within the marketing industry.

What Is Dentsu Japan’s New Strategy “AI for Growth 2.0”?

Satoru Yamamoto: In March 2025, Dentsu Japan announced a new strategy called “AI for Growth 2.0.” At its core is the belief that AI should go beyond improving efficiency to become a driver of growth and new value.  “AI for Growth 2.0” is built upon three main pillars.

The first pillar is the evolution of AI models. This means developing and strengthening proprietary AI models that leverage Dentsu Group’s marketing capabilities. Second is the evolution of AI marketing, which aims to generate entirely new marketing strategies and customer experiences by leveraging AI. The third pillar is the promotion of business transformation through AI, or “AI transformation.” For more details on this, our Executive Vice President Daisuke Kobayashi provides an in-depth explanation in another article, so I will focus here on the evolution of AI models and the evolution of marketing.

The Evolution of AI Models

We are currently developing AI models by taking two main approaches: the “People Model” and the “Creating Thinking Model.”

The People Model aims to simulate the overall Japanese population using AI. Dentsu Group has built a proprietary Large Language Model (LLM) based on data from a large-scale survey of 150,000 people. While this number may seem limited , by combining a range of diverse conditions, this model can be expanded to simulate 100 million Japanese people, including those outside the original sample.

Traditionally, research and analysis by humans would have required significant time and cost, but this model allows for near real-time, survey-like simulations.

The Creating Thinking Model teaches AI to think like agency creators, supporting advertising expressions and idea generation. The first version learned the thought processes of copywriters to generate emotionally engaging advertising copy, and the second version has expanded capabilities to  generate images as well.

Zoom

In this way, we are promoting the practical and highly accurate use of AI through the dual approaches of deeper consumer understanding (People Model) and the advancement of creative expression (Creating Thinking Model).

The Evolution of AI Marketing

What does it mean for marketing to evolve?

We believe that the essence of marketing in the AI era lies in allowing AI to handle daily routine tasks while enabling marketers to focus on their core creative duties—like inspiring people and making an impact on society.

At the core of this shift is the development of various products leveraging the AI models. The Dentsu Group has organized marketing operations into seven categories —Planning, Research, Creative, Journey & Media Planning, Execution, Experience, and Measurement & Optimization—and is developing and implementing products tailored to each of these categories.

We are also working on combining these products as conversational, AI agent-based solutions, making it possible to utilize them in an integrated way.

AI Agent Demo Video

Going forward, we will not only leverage assets owned by the Dentsu Group, but also utilize the unique data of each company to achieve highly accurate consumer simulations as ∞AI Customer Twin. Based on these simulations, we will design more tailored journey and media plans, offering them as even more practical and effective solutions.

Satoru Yamamoto (CAIO, Executive Officer)

∞AI Marketing Hub: Advancing and Streamlining Marketing Operations

Hironobu Konami: In day-to-day marketing operations, a variety of challenges arise—such as research, data utilization, human resource constraints, and the accuracy of decision-making. We believe that these issues can be addressed with AI agents.

For example, by using AI agents in research tasks, you can significantly shorten the time needed for surveys and analysis, making it possible to reallocate those resources to implementation and improvement of your strategies. Additionally, by teaching dispersed data to AI and embedding it into AI personas or agents, you can foster data-driven marketing that leverages accumulated know-how.

From a human resources perspective, implementing AI agents eliminates dependency on individual employees and promotes the standardization of workflows. Through performance forecasting by AI personas and agents, you can increase the likelihood of successful strategies and make more precise decisions.

Dentsu Digital’s ∞AI Marketing Hub comprehensively supports these diverse marketing challenges. This solution is composed of three main components.

At its core is the data-driven AI persona ∞AI Customer Twin, which is generated from each client’s proprietary data. This AI persona serves as the foundation, and is seamlessly connected with ∞AI MC Planning—responsible for media planning, execution, and measurement in the digital domain—and ∞AI CX Planning, which handles research, customer journey design, and comprehensive CX planning.

Zoom

By combining these products, we envision a world where advanced and efficient planning is made possible through a deep understanding of customers backed by data. Centered on ∞AI Customer Twin, ∞AI MC Planning and ∞AI CX Planning work in tandem, resolving many of the challenges previously faced in marketing operations. We believe this approach enables more strategic and efficient management across all marketing activities.


∞AI CX Planning: Covering Everything from Research to CX Design

∞AI CX Planning has three major features. First, by utilizing Dentsu Group’s large-scale consumer data (from 150,000 people) and social media data—combined with proprietary company data—it can generate highly realistic and accurate AI personas, dramatically increasing the “resolution” of customer understanding. Second, by aggregating the practical expertise in CX/UX planning that we have cultivated over the years into AI agents, it enables high-quality operations without relying on individual expertise. Third, by conducting research through AI personas, it overcomes traditional resource constraints and delivers rapid, high-frequency feedback.

Zoom

In the future, we aim to realize a multi-agent environment in which the CX Planning Agent serves as the central coordinator, autonomously executing all tasks in collaboration with the Research Agent and Journey Agent. Centered on ∞AI Customer Twin, we will continue to develop a new business infrastructure that enables seamless, automated workflow—from research to journey design—in a one-stop, end-to-end process.

Hironobu Konami (Head of Experience & Product Division)

∞AI MC Planning: Bringing Innovation to Media Planning

Koichi Sugimoto: ∞AI MC Planning is a solution that leverages AI technology to advance and streamline digital media planning operations.

This solution covers the entire process of standard digital media planning— Research, STP Analysis, Planning, Execution, and Analysis/PDCA (Plan-Do-Check-Act cycle)—and supports an optimized, end-to-end marketing process by combining three types of AI agents: strategic planning, media planning, and execution.

Zoom

In the strategic planning phase, the Strategic Planning Agent supports strategy design by leveraging Dentsu Group’s extensive expertise. Next, the Media Planning Agent provides end-to-end support—from budget allocation and target audience design to the creation of submission briefs, automatic submissions to media platforms, and campaign delivery. Furthermore, the Execution Agent integrates with the group’s dashboard “MIERO,” handling tasks such as proposing action plans and generating graphs based on delivery results data, as well as optimizing allocation strategies.

By integrating these three agents, ∞AI MC Planning aims to achieve two key advancements going forward. The first is enhancing functional capabilities, such as expanding targeting options and recommended media, improving simulation accuracy, and reinforcing document export features. The second goal is autonomous evolution: we are developing an “Agent Builder” that will allow for flexible design and construction of new agents as additional tasks arise in response to changes in the marketing environment.

To realize a co-creative work process between humans and AI, it’s important to review current workflows and implement and evaluate these solutions in a phased manner. After adoption, clearly defining goals, building the right framework, and promoting talent development in an integrated way are essential for integration into day-to-day operations.

Through ongoing, hands-on support, we aim to realize innovative work processes together with our client companies.

Koichi Sugimoto (Head of Strategy Division / Executive Director, Data & AI Division)

For inquiries regarding the marketing AI agents, please contact us here.

Please include “Attn: AI Hacked Marketing Office” in the inquiry details section.