Can dashboards become good storytellers?

Can GenAI overcome the limitations of data visualisation and help people make better choices in complex scenarios?

The problem with charts

When it comes to understanding complex data, traditional charts and dashboards can often feel like trying to see the world through a narrow lens. They’re inherently limited by the way humans perceive still images—we can only capture so much at a glance. This often leads to an oversimplification of the data, stripping away the rich context and nuances that are essential for truly informed decision-making. On the other hand, when charts try to pack in more information, they risk overwhelming us with complexity, making it difficult to discern actionable insights.

Imagine trying to piece together a detective story by only looking at a series of disconnected snapshots. Each image might tell a part of the story, but without the connecting details, the bigger picture remains elusive. This is precisely the challenge businesses face with traditional data visualization methods. These methods are limited not only by human cognitive constraints but also by the inability of machines to grasp the context behind the raw numbers. Machines can process vast amounts of data quickly, but they struggle with understanding the semantics—the why and how behind the numbers—because they lack the necessary context and meaning.

During our collaboration with data scientists, we uncovered how these limitations and biases can negatively impact critical tasks like demand forecasting. Charts and dashboards, while useful, often fail to capture the intricacies of real-world scenarios, leading to decisions based on incomplete or misunderstood information. Recognizing these gaps, we embarked on a journey to design a new system that goes beyond traditional visualization. By introducing the concept of perspective, we aimed to create a more comprehensive and nuanced view of data that helps bridge the gap between raw information and meaningful insights.

Meet ConAgra brands

Meet ConAgra Brands, a household name known for filling kitchens with favorite food products around the world. With operations spanning both large and small retailers globally, the precision of ConAgra’s supply chain is crucial. It’s not just about keeping shelves stocked—it’s about affecting their market share, the success of brand promotions, and ultimately, their bottom line. Accuracy in forecasting is vital for sustainability and investor confidence. To tackle this, ConAgra partnered with Accenture Applied Intelligence for the tech solutions, while Momentum Design, a leading UX agency, was brought in to get a handle on ConAgra’s needs and co-design the way forward.

Objectives

This initiative was far more than just developing another dashboard—it was about fundamentally transforming ConAgra's operations to achieve significant commercial and operational gains. At the core of this transformation were the limitations of traditional charts and dashboards, which could not fully support ConAgra’s ambitions for optimizing supply chain management through improved forecast accuracy and faster decision-making. Our key objectives were:

1. Automate and Innovate: Traditional manual processes were slowing down ConAgra’s ability to respond to market changes effectively. By shifting from these manual tasks to advanced technologies like machine learning and data science, we aimed to streamline operations and improve overall efficiency. This automation would allow ConAgra to handle more data with greater speed and accuracy, providing the insights needed to stay ahead in a competitive market.

2. Redefine Roles and Responsibilities: With the introduction of new technologies, there was a need to realign organisational roles and responsibilities. This meant changing how people within ConAgra worked, integrating new tools and workflows that aligned with emerging goals and expectations. The goal was to foster a more agile and responsive workforce that could leverage data insights to drive strategic decision-making.

3. Enhance Decision-Making Quality: One of the most critical objectives was to improve the quality of decision-making processes across the supply chain. This required moving beyond the limitations of basic charts, which often provide only a superficial view of the data. By designing advanced, intuitive dashboards, we sought to present data in a way that tells the full story, offering deeper insights and greater context. This approach would empower leaders at ConAgra to make quicker, more informed decisions, directly impacting the company’s ability to optimise its supply chain, improve forecast accuracy, and ultimately drive profitability.

At the heart of this transformation was a need to rethink how data is visualised and used. The goal was not just to present data but to provide a holistic view that integrates operational and strategic insights, helping ConAgra achieve their ambition of optimised supply chain management and sustainable commercial success.

Discovery phase: uncovering the limitations

As we began working on the project, it quickly became apparent that traditional data visualisation methods had significant limitations, particularly in the context of ConAgra's complex supply chain forecasting needs. Here’s what we uncovered:

1. Oversimplification and misinterpretation: Standard 2D charts, while commonly used, can make complex data appear deceptively simple. This oversimplification often leads to misunderstandings or misinterpretations, akin to trying to understand a city’s layout by looking at a basic, unlabeled map. Important details and nuances are lost, which can result in decision-makers missing critical insights.

2. Lack of context and interactivity: While adding a third dimension to charts might seem like a solution for depth, it often introduces more confusion than clarity. 3D visuals can be harder to interpret correctly, especially without interactivity that allows users to dig deeper. What is needed is a way to present multi-dimensional data that users can engage with intuitively, providing both clarity and depth.

3. User-centred design limitations: Effective data visualisation isn't just about presenting numbers; it's about understanding the users—what they need, what questions they’re trying to answer, and how they interact with the data. Without a user-centred design approach, visualisations fail to tell a compelling story, leaving users with numbers that lack meaning and context.

Approach: reframing the problem

Faced with these challenges, we set out to rethink our approach to data visualisation, particularly focusing on complex scenarios like supply chain forecasting for ConAgra. Our strategy needed to enable quick adaptation to changing market conditions, provide a comprehensive view of the data, and facilitate faster, more informed decision-making. This meant moving beyond traditional static visuals to a more dynamic and interactive design that could tell the whole story.

Key issues: understanding the gaps

1. Limited visualisation capabilities: Traditional charts and graphs are effective at showing what happened, but they fall short in explaining the why and how. For instance, a CFO reviewing a simple bar chart might see sales numbers falling but lack insights into the factors driving the decline, such as market trends or supply chain disruptions.

2. Complex scenarios need complex narratives: In real-world business environments, data is rarely isolated. Multiple datasets interact, creating a complex narrative that a single chart cannot capture. It’s like trying to represent a full orchestral piece using only a single instrument—it lacks the richness and completeness needed to understand the whole picture.

3. Avoiding overcrowded visuals: Attempting to convey too much information in a single visual leads to clutter and confusion. Overcrowded charts fail to provide clarity, overwhelming users and making it harder to extract actionable insights. Our goal was to create visuals that were detailed yet straightforward, offering clarity without sacrificing depth.

Design phase: crafting the solution

To address these challenges, we focused on designing dashboards that did more than just present data; they offered interactive experiences that facilitated deeper understanding and exploration:

1. Interactive dashboards: We developed dashboards that allowed users to manipulate and explore data from multiple perspectives. By enabling users to view data from different angles and dive into specific areas of interest, these dashboards provided insights that static charts simply couldn't deliver. This interactivity meant users could better understand the relationships between different data points and uncover hidden insights.

2. Standardising best practices: To maintain consistency and reliability across all visualisations, we created a comprehensive library of visualisation patterns. This library served as a standard reference for designing effective and intuitive data visualisations, ensuring that all reports and dashboards followed best practices. Consistency in visual design helps users navigate and trust the information presented, making decision-making more efficient.

3. Innovative visualisation techniques: We introduced new methods for displaying complex data, incorporating interactive elements that allowed users to ask questions directly within the dashboard. These features enabled users to receive real-time feedback and insights, reducing the need for back-and-forth communication and significantly speeding up the decision-making process. By focusing on the narrative aspect of data, we helped users move from raw data to actionable insights quickly.