“We see people acting, but the constraints under which they are acting are largely invisible to us.”
Professors Jeffrey Pfeffer and Robert I. Sutton, “Hard Facts, Dangerous Half-Truths & Total Nonsense: Profiting from Evidence-Based Management” (2006)

In my analytics work and leadership, I use the above quote as a reminder that data can only reach so far into the reality of the world. Analytics relies on data as a record and evidence of actions. However, achieving clarity requires recognizing what data alone cannot capture and interpreting the potential gaps. Sometimes we know what we cannot record, sometimes we do not. Other times we can observe the actions that data miss and then use experience and intuition to provide interpolations and extrapolations beyond the data. Sometimes the data escape us because the observed actions are rooted in lessons learned from past actions that are lost to the fog of time. As a result, data is a representation of human reality and not itself human reality. Analytics offers one kind of perspective for interpreting how the data reveal insights about reality.

Data for Joy

These sobering limitations inherit extra significance and meaning in “Data for Joy: Restoring Balance in the Age of Algorithms” by Phylis Savari. In Data for Joy, Savari relates how Plato’s allegory of the cave cautions data professionals to understand and appreciate how data-driven processes generate projections of reality (pages 134-137). Savari’s illustration of human projection in the data-driven process provides another layer for understanding how the Insights Supply Chain works to bring clarity to data and transform it into business value.

According to Savari, “each projection [offers] the opportunity for personal values, biases, and mistakes to be infused.” In the context of  Savari’s experience with transactional data, “engineers used technology…to project a consumer’s experience into transactions; data analysts projected these transactions into business success metrics, insights, and stories; and decision-makers projected these success metrics, insights, and stories into action items to help them achieve their business goals.” Thus, like Plato’s cave, each step in the Insights Supply Chain may project the shadows of objects. Each step in the chain observes shadows as the reality for generating the next projection.

The Data Org Matrix (DOM): Insights Beyond the Shadows

The significance of the projection problem depends on how a company organizes its data professionals. The amount and intensity of knowledge-sharing and collaboration must be strategically aligned with the distribution of knowledge and people. The Data Org Matrix (DOM) in the Insights Supply Chain provides the framework for understanding this strategic principle.

Each of the four sections of the grid describes the kind of collaboration and knowledge-sharing required to most effectively transform data into insights aligned with business objectives. Along the way, a well-organized data ecosystem will minimize the dysfunctional impact of projecting shadows as reality. Let’s explore two examples, each with a different and separate organizational context. The first context exists in the upper left quadrant of the generalized data professionals, centrally organized. The second context exists in the lower right quadrant of the specialized data professionals working in a decentralized organization.

Generalized and Centralized

In the upper left quadrant, data professionals have generalized knowledge, that is few people are deep experts in any one data domain and data professionals are centralized in one data organization. Thus, according to the Data Org Matrix, leadership must focus on empowering individuals to drive knowledge-sharing through tools like shared dashboards or working groups. While centralization facilitates collaboration (through more closely aligned priorities, objectives, and success metrics), generalization disperses the knowledge required to move data through the Insights Supply Chain. For any given project or task, a different set of individuals may be required to gain understanding beyond the projections.

The hapless prisoners of Plato’s cave are forced to stare straight at the shadows as their only reference point for reality. Taking some license with the allegory to apply it to the upper left quadrant and expand beyond Plato’s description of an individual journey of enlightenment, think of these prisoners as generalists who manage to free themselves at similar times. Once in the light (Plato’s sun is the “single source of truth”), these generalists share their individual insights to bring clarity to the objects in the cave, which they once only understood in shadow form. In the upper left quadrant of the Data Org Matrix, data professionals look locally and seek out individuals to construct insights beyond the projections and the shadows.

Specialized and Decentralized

In the lower right quadrant, specialized data professionals organized in a decentralized manner must rely on teams to drive collaboration. While specialization facilitates the identification of experts needed for a given project or task, the dispersion of the teams across the organization make collaboration more challenging (think competing priorities and demands). Thus, leaders must focus on clearing cross-functional, cross-organizational paths for collaboration and empowering teams of experts to develop effective workflows.

In this second organizational context, think of the prisoners in Plato’s cave as experts who are unaware of their specializations until they find their way into the light of Plato’s sun. Once in the light of the single source of truth, they find it easier to strengthen their insights with others who share their (now discovered) expertise. Unlike the generalists who share knowledge to develop integrated insights, the specialists must collaborate on a workflow that assigns clear roles in the Insights Supply Chain to achieve clarity and piece together domain-specific knowledge into integrated insights. In the lower right quadrant of the Data Org Matrix, specialists may need to create parallel organizational structures like cross-functional task forces to work through conflicting objectives and to remove organizational barriers to collaboration.

Do AI Agents Stare At Shadows?

In “The Case for Human Conversations: Why Data Professionals Remain Critical in the Age of AI Agents Like Gemini in Looker“, I critiqued Google’s implication that its AI Agent in Looker could free business users of their dependence on data professionals. Plato’s cave provides another poignant perspective. These AI Agents may provide the next version of staring at the shadows of objects. Without conferral with data professionals, the business user could mistake better understanding of shadows for the spotlight on the object of interest. Even the AI Agent must be fully integrated into some component of the Data Org Matrix. Based on my experience, I can see a future of organizations getting tripped up by business users running with AI-generated insights devoid of the context, nuance, and experience delivered by data professionals.

Step Away from the Shadows

Whether your organization is trying to develop a data strategy aligned with business objectives or taking the next step to create an AI Strategy well-integrated into your data strategy, Ahan Analytics can help. The Insights Supply Chain will transform the way your organization organizes its data professionals, providing clarity and helping the entire data ecosystem step away from the shadows and into the light. Contact Ahan Analytics, LLC for more details!

Questions? Need more clarity? Feel free to use the comment section below!

---