By Jeremy Katz, Chief Product & Solutions Officer, Data & Identity, denstu
Michael Fuchs, Head of Strategy & Planning, Data & Technology, dentsu
The era of static, monolithic systems is over. As AI-driven agents begin to make decisions and manage actions, brands that cling to rigid architectures will be left behind. The only way forward is composability: a modular, adaptive approach that lets you plug in new tools, pivot fast, and evolve in real time. It’s the strategic foundation of an agentic future. In this article, we’ll cover the characteristics to incorporate into your strategy and the benefits that brands that embrace composability will see.
For years, organizations have relied on monolithic, one-size-fits-all systems. They worked well enough when digital experiences and customer expectations moved at a slower pace. But they can’t keep up with the speed, scale, and complexity of what comes next: an agentic future powered by AI-driven decision-making.
AI agents, whether customer-facing or behind the scenes, are built to act, adapt, and improve. Their potential is limited if they’re forced to operate within rigid systems that can’t change as quickly as they do.
To truly unlock the promise of intelligent automation, brands need to embrace composability as their strategic foundation: a modular, plug-and-play approach that makes it possible to experiment, integrate, and evolve in real time.
What the Agentic Future Looks Like
AI agents make decisions and take action on your behalf within defined parameters and oversights. Unlike traditional bots, intelligent agents use context, learn from outcomes, and adapt, and in an agentic future they’ll work together agent-to-agent (A2A).
But here's what most conversations about agentic AI miss: the most consequential interactions won't just happen inside your organization, they'll happen at its edges. When your agents meet external agents. When your brand comes face-to-face with a consumer's AI assistant that's evaluating options, comparing prices, and making purchase decisions. We call this Agentic Diplomacy: the emerging discipline of designing, deploying, and governing AI that doesn't just work for you internally, but represents your brand in the external ecosystem. This is where the next competitive advantage will be won or lost.
A2A in action: Consider an abandoned cart. Traditionally, a generic 'you forgot something' email arrives 24 hours later. With an agentic approach, specialized agents act within minutes, analyzing behavior, checking inventory, evaluating a targeted discount, and delivering a personalized message. Now take it further: that same shopper has an AI assistant comparing your brand against competitors in real time. Your agents aren't just talking to your internal systems anymore, they're representing your brand to another agent.
Here’s the catch: agents can’t deliver this level of value if they’re not in an environment that’s as adaptive as they are. That’s where composability comes in.
Composability as the Foundation
Agents thrive on modularity and need flexibility to survive. They may need to plug into new tools, swap out components, and orchestrate across domains.
Composability provides:
- Dynamic integration: Agents can pull the right data and tools they need when they need them.
- Continuous evolution: Modular systems make it easy to update, swap, or improve components as technology evolves.
- Cross-functional interoperability: Agents can work across teams to connect insights and findings across marketing, sales, service, etc.
In an era where AI capabilities double every few months, closed platforms can become obsolete before implementation is even completed. By having a composable architecture, organizations can have seamless implementation of current and future technologies, integrate data, and provide the foundation for agentic workflows.
How to Get Started
Moving to composability is more straightforward than it sounds. The main challenge isn't the technology, it's the semantic layer: ensuring all your systems speak the same language. At a base level, this can be solved with metadata, good taxonomies and clear data standards that give agents the context they need to act intelligently. The effort required depends on how far your organization already is with data governance.
From there, take a strategic approach to make it work for you today and for future goals and initiatives:
- Define use cases: Where will agentic workflows deliver the most value? Start specific, not sprawling.
- Align your organization: Get cross-functional buy-in. Composability affects many parts of the business, not just IT.
- Decide on build vs. buy: Some components you'll develop internally, others you'll integrate from best-of-class vendors.
- Legal and compliance review: Agents making autonomous decisions means defining clear guardrails.
- Develop a proof of concept: Test the approach with a contained use case to ensure responsible scaling.
- Scale and iterate: Learn from what works, adjust what doesn't, expand gradually.
The fundamentals haven’t changed. What’s changed is the scale of the opportunity and the cost of getting the architecture wrong. With composability, you're not betting your entire transformation on getting everything perfect in one massive implementation. You're building iteratively, learning as you go, and maintaining the flexibility to pivot as both technology and business needs evolve.
The Bottom Line
The agentic future isn't a far-off scenario. It's happening now, and the architecture decisions you make today will determine whether your brand is leading the charge or trying to catch up.
dentsu.Connect is built to make this real. It’s an intelligent operating system that unifies the entire marketing lifecycle, orchestrates work end-to-end, and augments every decision with human governance. Learn more about it here. The question is how quickly you can get started, and with the right team and partners.
