Designing a new agentic collaborative workforce

In the world of marketing, staying ahead of the curve is more critical than ever. As leaders at Accenture, we are excited to share our journey of reinventing our Marketing function using AI agents. This journey is leading us to cut the steps in running the average campaign from 135 to 85 and get us to market 25 – 35% faster. And the use cases for broader outcomes are creating themselves every day. Because reinvention using AI Agents is more than technology. It is about people, process, technology and data, and a deep understanding of entirely new ways to work and agentic AI’s role in that future state; it's about integrating these agents as valued assistants and team members to aid marketers who deeply value curiosity and seek to dig deep to find the right marketing and communications path forward. And it’s about taking on some new questions at the interface between our people and this new technology. Here’s the story of our experience driving the reinvention of Marketing, with AI agents by our side - we are sharing our story to help our clients better succeed at theirs.

Jill Kramer: The Marketing Perspective

As the CMCO at Accenture, I've always been passionate about finding new ways to drive success, creativity and performance. Recently, we embarked on a journey to reinvent our Marketing function using AI agents. This was more than technology; it was a mission to drive a cultural shift that required a deep understanding of how AI can enhance our team's capabilities and allow us to work in new ways.

1. Identifying the right work: We started by identifying the steps within a marketer workflow suitable for automation and augmentation. Market research, content creation, and data analysis were prime candidates identified by our marketers. By measuring current workloads and taking on pain points, we designed new processes that leveraged the strengths of both humans and agents. We clearly outlined what new value we could create if we reimagined the way we worked.
   
   
2. Creating our ideal new reality: With a clear idea of how our work and processes could be reimagined, we allowed ourselves to define new business outcomes, challenge our levels of productivity and performance, and, maybe most importantly, create a strong belief in our unique and critical role in bringing valuable insight back to the business. Steeped in client insights, informed by the reality of our competitive environment and determined to bring forward anything but the “normal” way we had “always” driven success in the past, we could and would be a fundamentally different function.
   
   
3. Managing uncertainty: We agreed that we would embrace what we didn’t know, agree on acceptable and responsible guardrails of experimentation and create a culture of exploration free of false certainty. We are a group of experts, and we needed to trust our instincts. And we did. Key to the ability to do that was building in a flywheel of learning and feedback. It is easier to venture into the unknown if we felt confident that the feedback would be timely and 2-way – agent to marketer and marketer to agent.
   
   
4. Preparing our data: We needed to lean into the state of our data – all the way in. This was a reality the leadership team needed to own, not freelance out to the data team. We needed to build a solid data foundation, integrate and enhance our data and ensure we had advanced analytics and AI to deliver strong performance. Knowing this data deeply was the only way we could design the optimal experience for our people, by ensuring their agents were operating in the right data set to bring valuable information and suggestions to the team.

And we knew we had partners within Accenture to assist us in making this vision a reality.

Karalee Close: The Talent and Organizations Perspective

As the Head of Talent & Organization at Accenture, I am deeply involved in how we redesign work and workforce to support people and intelligent agents in working together effectively. The technology is exciting and yet there are still many questions about how all of this will all work in practice – in a real organization and beyond a ‘proof of concept.’ We’ve been working with leading clients and our own Corporate Functions to define the steps, capabilities and practices that meaningfully ‘bring the agents to life’ at pace and at scale. We’ve defined a structured and repeatable approach, which we’ve tested and refined by working within our own Corporate Functions. In practice, we’ve been helping Jill, and her Marketing team, effectively lead in the context of uncertainty, to be architects of change, and to enable marketing teams and intelligent agents work together more effectively. We’ve been tackling the deeper questions on skills – using analytical approaches to better understand the current and future skills as well as how we can develop future skills more quickly through real time and personalized learning. We’ve also been tackling how we are adapting the organization and operating model and delivering continuous change as both people and agents learn. We’ve had to find new and structured ways to create human + AI connection in the work and we believe these learnings are incredibly important to all organizations.

1. Inspiring change: We thought deeply about the ‘why’ – tackling the issues of ‘will these things take our jobs head on.’ We aligned our messaging to our core purpose that resonated with marketing teams – to be the best Marketing function in the world. Yes, we also wanted productivity but there needed to be a bigger mission to really bring creativity and engagement. We used influencer networks, 2-way communications and data on sentiment with behavioral science techniques to tailor messaging and address the factors that really developed trust and commitment.
   
   
2. Onboarding and integrating the agents: We used a structured onboarding process to help AI agents make the right connections and provide feedback to improve their performance. We borrowed from our HR best practices on team integration to ensure agents were introduced to the Marketing team and understood their roles. We also created feedback loops that enabled simple, 2-way feedback between marketers and agents, which turned out to be incredibly important in establishing collaboration and mutual learning.
   
   
3. Enabling dynamic skills and learning: We used new tools to understand current and future skills. We established training experiences and feedback mechanisms to ensure that AI agents were well-equipped to handle their tasks and that our marketers developed new skills in parallel. We focused on role-specific training, which tailored the approach to specific roles played by the agents. We also emphasized continuous learning for both humans and agents. Learning in the flow of work and personalization was critical – from basic AI fluency and experience to build trust to tackling deeper questions on how to get the best from the agents and approaches to responsible AI. Our LearnVantage platforms made this possible and allowed us to capture meaningful data that helped us learn along the way.
   
   
4. Managing a new organization: Redesigning the work meant redesigning jobs and a new level of performance management. We re-visited the organization structure and blueprint to better organize the remaining work and the changes in jobs. We also needed to manage the performance of the intelligent agents. Line managers were tasked with overseeing both human and agent performance. We needed to establish new metrics to measure the performance of both humans and agents as the work evolved. We are also considering the possibility of performance cycles for agents to recognize their contributions as well as the support provided by their human mentors. As the work of people and intelligent agents continues to change, we will continue to evolve and optimize the organization. As our use of Intelligent agents expands in Marketing and across Accenture, we also expect to come to some deeper questions on organization including challenging traditional boundaries between functions and finding new ways to manage the capabilities of multi-agent organizations.

Lan Guan: The AI Perspective

As Accenture’s Chief AI Officer and the Head of our Global Center for Advanced AI, I oversee our comprehensive AI capabilities—both internally at Accenture and externally in the market. I help clients navigate their most pressing AI challenges. For this project, my role was to lead the transformation of Accenture’s Marketing function using Accenture’s AI Refinery platform with autonomous agents to help create and run smarter campaigns faster. This initiative demonstrated the power of AI-driven solutions in enabling reinvention and delivering tangible value. From this experience, three critical lessons emerge:

1. Shifting from rigid processes to dynamic workflows: Many companies have rigid business processes spanning multiple systems, which makes reinvention with AI cumbersome. Our marketing function is no exception. This is because traditional software systems powering these business processes are instruction-based, relying on pre-programmed business logic. In the new model, organizations must shift their thinking toward process reinvention and dynamic workflows, where not everything is pre-programmed or rigidly defined. An intention-based architecture like this is based on AI agents. Unlike preprogrammed systems, these architectures are designed to dynamically allocate resources, integrate data across silos, and make decisions based on evolving goals or "intentions." This flexibility enables AI to respond intelligently to complex, real-time scenarios, and optimizing workflows.
   
   
2. Engaging Marketers in training AI models: Marketers play a pivotal role in providing feedback and actively helping train AI models. This collaboration ensures that the AI systems become more attuned to the unique needs of the business, resulting in smarter and more effective campaigns over time. Our AI agents were created to help perform market research, develop new content for marketing briefs, perform strategic planning, and execute marketing plans. Our marketers provided the human insight and reviews on first drafts developed by the AI agents, refined the content, and made real-time adjustments as needed.
   
   
3. Breaking silos for strategic decision-making: Cross-functional collaboration between humans and AI agents is essential for breaking down traditional silos, empowering marketers to make more informed and strategic decisions. The true potential of AI is not just a technological advancement—it lies in reimagining workflows and fostering seamless integration between humans and AI agents. This potential was fully realized during the design and implementation of our AI-powered marketing workflows. Marketers worked closely with AI research scientists, contributing their domain expertise to shape the system and ensure its relevance. This partnership created a feedback loop of continuous improvement, where both the technology and processes evolved in tandem. This approach—marketers collaborating side by side with AI experts—represents the "holy grail" of human-AI collaboration, enabling organizations to achieve strategic reinvention and unlock lasting value.

Conclusion

Reinventing our Marketing function with AI agents has been a transformative journey – and a collaborative effort across Marketing, Talent, and our AI teams. By treating these agents as valued team members dedicated to elevating the work of our ingenious humans, we’ve not only delivered breakthrough performance but also fostered a culture of curiosity and innovation. The future of marketing is here, and it’s a blend of human creativity and AI intelligence. Join us on this exciting journey as we continue to explore the endless possibilities of AI in marketing and operations.