Dentsu democratizes analytics with Fabric, delivering 55% faster data replication

Across the advertising and media industry, data is growing faster than teams can keep up. Campaigns span continents, privacy rules shift constantly, and clients expect real-time results. But behind the scenes, this creates significant complexity.

At Dentsu, like many organizations, the challenge was more than a systems issue. Teams across global offices were spending increasing amounts of time prepping data before they could begin to analyze it.

A hybrid mesh to balance governance and regional autonomy

Dentsu’s data infrastructure had reached a tipping point. Silos across regions, inconsistent tooling, and slow reporting cycles were making it difficult to deliver timely insights. The leadership team initiated a full rebuild of the enterprise data architecture to address these issues. The objective was to create a unified, scalable platform that could support real-time analytics, streamline collaboration, and prepare the organization for AI integration without adding complexity.

The company had already built a strong foundation with Microsoft cloud services, including the Microsoft Dynamics 365 suite for enterprise resource planning (ERP), Power BI, and Microsoft Entra ID for identity and security. The team wanted to evolve beyond fragmented analytics and move toward a unified, enterprise-grade platform. Dentsu selected Microsoft Fabric to streamline data workflows, reduce reliance on point solutions, and support scalable, AI-ready analytics. By implementing the semantic layer in Fabric, Dentsu established centralized gold-layer datasets and standardized KPIs, while enabling regional teams to customize insights for local needs.

Patrick Sura, Global Reporting Architect at Dentsu, states, “Fabric has evolved beyond self-service analytics to a full enterprise-class platform. By combining data warehousing, lake storage, and Power BI, we’ve unlocked a foundation that sparks curiosity and innovation. People naturally ask, ‘What can I build on top of this?’”

Microsoft Purview will soon reinforce governance and more secure data access, and Power BI templates have helped maintain consistency across markets.

Engineering a modern, Microsoft optimized data architecture

In partnership with oneLake GmbH, Dentsu accelerated its migration to Fabric, applying deep technical expertise to design a scalable, enterprise-grade analytics platform. Dentsu then built a domain-based architecture and medallion layers on Microsoft Azure to improve data sharing and concurrency across global teams.

Dentsu enhanced its Azure Synapse Analytics platform by adopting domain-based architecture and medallion layers to improve data sharing and concurrency across global teams. Raw data from core enterprise systems such as Dynamics 365 ERP and SAP is ingested through the company’s existing data pipelines into Azure storage. There, it serves as the foundation for downstream processing and analytics in Fabric. 

Using Apache Spark pipelines in Fabric, Dentsu then processed these inputs at scale, normalizing diverse formats, enriching data, and preparing it for high-performance querying. Central to the new estate, Fabric Data Warehouse consolidated structured datasets, enabling efficient cross-domain analytics while minimizing duplication. 

Power BI played a critical role by delivering curated, role-based dashboards and semantic models tailored for executive, finance, and regional teams. Meanwhile, data ingestion is orchestrated within the data engineering environment of Fabric, enabling reliable, high throughput data flows between source systems and downstream analytical layers.

Ebad Uddin, Director, Enterprise Data & Analytics at Dentsu emphasizes, “We’re not just building another set of dashboards. We’re creating a real-time, enterprise-wide decision engine. Microsoft Fabric helps us unify data, break down silos, and deliver fast, trusted analytics that empower every team member.”

Tackling migration challenges with a phased approach

Dentsu piloted the migration with one domain as a prototype to stress-test capacity sizing, workspace distribution, and governance models under real-world conditions, addressing challenges early.

Key technical components included the Fabric Migration Assistant, which streamlined dataset transitions from existing environments, and FastTrack assimilation for Dynamics 365, which dramatically reduced data replication time from over 45 minutes to under 20 minutes, unlocking near real-time updates for critical campaign-level metrics. Additionally, automated workspace provisioning scripts helped ensure that permissions and security policies were aligned and in place from day one.

Dennis Mausbach, Solution Architect & Co-Founder at oneLake GmbH, reflects, “With extensive testing, we experienced all of the issues, but that helped us mature quickly and educate the rest of the organization.” Weekly knowledge transfer and open-door sessions fostered collaboration and rapid problem-solving.

Unprecedented performance gains

Within months, data replication speeds improved by 55%, delivering near real-time insights and boosting analyst productivity. 

Consolidating workloads into the Fabric environment also lowered licensing and maintenance costs, enabling Dentsu to support enterprise performance management (EPM) reporting with a comprehensive, 360-degree view of agency network performance. 

As Sura explains, “We help our management see the performance of our agency network in a 360-degree view. It’s a good old controlling exercise, transformed into a cloud-based, modern way of working.”

Preparing the Fabric estate for generative AI 

With a unified data foundation established, Dentsu is advancing its analytics capabilities by assimilating generative AI. It seeks to empower leaders across global offices—from Tokyo to Toronto—to ask strategic questions in their native languages, while Copilot in Fabric understands, interprets, and instantly delivers clear visualized answers.

Sura explains how this works in practice: “For example, the core analytics model is maintained in English, but a leader in Tokyo can ask a question in Japanese and receive a fully visualized response. The magic lies in the ability of Copilot to translate the query, interact with the model, and return actionable insights seamlessly.”

Achieving this requires meticulous preparation, including cleaning and structuring decades of historical data. Semantic models, well-defined KPIs, and enriched metadata are essential to support the context-aware querying capabilities in Copilot. Looking ahead, Dentsu’s roadmap includes expanding use of predictive and prescriptive analytics to help stakeholders anticipate and evaluate the outcomes of decisions before acting.

Building a visionary future with unified, AI-ready analytics

Dentsu’s transformation is a cultural one, galvanizing teams around data-driven innovation. Internal communications highlighting success stories, recognition awards, and technical showcases on the new analytics platform. Events like FabCon Vienna gave internal champions a global stage to share lessons learned.

Sura notes, “We’re seeing strong momentum as regional teams bring forward new requests to build on this platform. That’s the real sign that transformation is taking hold and driving innovation.”

In a few months, as Dentsu completes its full migration to Fabric, the company is poised to reap immediate strategic benefits. These include reduced operational risk, enhanced agility, and a unified analytics fabric that forms a foundation for innovation. This platform is ready to incorporate next-generation AI capabilities, opening pathways for intelligent decision-making and transformational business impact. Uddin concludes, “We’re building a blueprint for the future. Unified, scalable, and ready for whatever comes next.”

Discover more about Dentsu on Facebook, Instagram, LinkedIn, X/Twitter, and YouTube.