Local governments today are under increasing pressure to be “data-driven.” When something goes wrong or needs detailed investigation, the response is almost automatic: gather more data, produce more reports, and build more dashboards. We tell ourselves that the more we collect, the closer we will come to the truth. Increasingly, we also rely on artificial intelligence (AI) not just to process data, but to read it for us, summarize it, and explain what it means. A useful starting point is to think in terms of being “data-informed,” rather than strictly “data-driven.”

The Conceptual Meditations series—through the dialogue between Dr. Concepto and Mr. Unhyde—explores the dual nature of the human mind and disciplined thinking: one that constructs meaning, and one that questions and tests it.

Read the first series entry, Why Conceptual Clarity Is a Hidden Building Block of Strategy. This second meditation examines a central issue: the challenge in modern decision-making is not a lack of data or information, but how we approach and treat it.

The Data-Driven Problem and the Data-Informed Approach

Mr. Unhyde: I keep hearing the term “data-driven”: data-driven decision-making, data-driven problem-solving. Are we really driven by data in everything we decide?

Dr. Concepto: Yes, but only to a certain extent.

Mr. Unhyde: That sounds ambiguous. Are data driving decisions or not?

Dr. Concepto: Data can certainly tell a story, reveal patterns, and highlight trends—especially quantitative data. Qualitative data, in turn, helps describe context and meaning, even when it cannot be easily measured. In both cases, data does reflect aspects of reality.

Mr. Unhyde: Then, as a kind of “reality sensor,” shouldn’t data drive decisions?

Dr. Concepto: Not necessarily. The term “data-driven” suggests that data itself acts as the driver—as if it were the pilot of decision-making. That implication is misleading. Data is not an actor. It does not think, interpret, or decide. A truly data-driven approach works best in automated environments. For example, a system may be designed to trigger an alert when a 911 call involves injury or death. In such cases, actions are predetermined and activated when data meets specific criteria.

Mr. Unhyde: Then why do leaders use “data-driven” when referring to complex decisions and processes like strategic planning?

Dr. Concepto: Because they want to signal that their decisions are grounded in reality and rely on evidence rather than intuition. The intent is valid, but the language oversimplifies reality.

Mr. Unhyde: So what would you suggest instead?

Dr. Concepto: “Data-informed” is gaining popularity. This term reflects that data contributes to decisions, but does not entirely control or predetermine them. Decisions also incorporate judgment, experience, stakeholder perspectives, and context. This approach requires interpretation, assumptions, and deliberate thinking.

Mr. Unhyde: So “data-informed” is more accurate?

Dr. Concepto: Correct. These terms represent fundamentally different approaches and thus must not be used interchangeably. “Data-driven” fits automated systems. “Data-informed” better describes human decision-making.

Data as a Universal Currency

Mr. Unhyde: It seems data has become a universal currency.

Dr. Concepto: Yes, but it is not a universal language of understanding and meaning. We live in the most data-saturated environment in history, yet disagreement has not diminished. Two people can examine the same data and reach entirely different conclusions.

Mr. Unhyde: And that happens inside organizations as well.

Dr. Concepto: Constantly. I have seen identical datasets support opposing courses of action, depending on what response individuals believe those numbers should trigger. In some cases, data is even selected to support a pre-existing conclusion rather than to test it. Data appears objective, but it is always interpreted through conceptual lenses—both personal (beliefs, knowledge, preferences, etc.) and environment-related (stakeholders' preferences, community expectations, etc.) When those conceptual lenses differ across teams and decision-makers, organizations end up debating interpretations instead of addressing underlying realities.

The Hidden Player: Thinking Through Concepts

Mr. Unhyde: So, the issue could not be the data or information themselves?

Dr. Concepto: Often, it is how we interpret it. Every chart, report, or dashboard requires interpretation, and interpretation happens through concepts. Concepts are mental constructions that help us understand the world. When those concepts are unclear or inconsistent, they create an illusion of shared understanding. In other words, conceptual clarity is key.

Mr. Unhyde: Can you give an example?

Dr. Concepto: During a job interview for a supervisory position in local government, I was asked if I was “good at multitasking.” I said yes, based on my experience as a military officer. English is not my mother tongue, and in my native context, “to task” typically means assigning work. To me, “multitasking” meant coordinating multiple efforts simultaneously - a managerial skill. But the interviewer meant something else: performing multiple tasks simultaneously as an individual—the same word—two different concepts. The root term was the same, but the overarching meaning had effectively reversed direction. Even today, this language-specific shift strikes me as conceptually inconsistent. From an analytical standpoint, increasing scale—from “tasking” to “multitasking”—should expand scope, not invert meaning. In my view, it is analogous to changing a noun from singular to plural and, in doing so, unexpectedly altering its fundamental category, i.e., changing its grammatical gender. This experience, though conceptual in nature, highlights a broader organizational issue: misunderstandings are often subtle. They can be professional, well-intentioned, and embedded in familiar language that appears self-evident until it is applied in practice. 

Mr. Unhyde: And this applies to data as well?

Dr. Concepto: Absolutely. Data does not speak for itself. Meaning is assigned through interpretation, and interpretation requires time and thinking effort. Modern organizations often prioritize speed over conceptual clarity. As a result, data collection accelerates while understanding lags behind. The real risk is not just fast decisions, but decisions based on assumed, rather than shared meaning.

AI as a Game-Changer and a Risk

Mr. Unhyde: How does AI fit into this?

Dr. Concepto: AI transforms how we handle data and information. It retrieves, summarizes, and even interprets information at extraordinary speed. Human thinking operates through two systems: fast (intuitive) and slow (analytical). AI excels at tasks that resemble fast thinking—rapid pattern recognition and summarization.

Mr. Unhyde: Is there a risk in that?

Dr. Concepto: Yes - the risk is outsourcing, meaning itself. There is evidence that AI can produce smooth, confident language without any real reference to meaning. The smoother the output, the easier it becomes to mistake it for genuine insight. Interpretation—assigning meaning—remains a human responsibility. It requires judgment, context, and reflection.

What This Means for Local Government Leaders

Mr. Unhyde: What should leaders do differently?

Dr. Concepto: Several practical steps can help ensure a more comprehensive and shared understanding of problems, whether they are brought by conceptual dichotomy, data, or information:

• Clarify meaning before planning or measuring. Ensure that key terms are consistently understood across departments before building metrics around them. In public safety, for example, call volume (aka calls for service) is the most commonly used metric. However, when you ask different professionals about their department’s call volume, answers may vary. Some report incident volume—the number of distinct emergency events; while others report response volume, which counts all units dispatched to those incidents. The latter can be two to three times higher than the former, making it essential to define terms clearly.
   
• Use analysts as facilitators of understanding, not just producers of data. Analysts add value not only through data production but by helping interpret information and surface underlying assumptions. 
   
• Question interpretations, not just data. Ask not only whether the numbers are correct, but how the data was collected, filtered, and structured. What assumptions shaped its interpretation, if relevant? 
   
• Allow time for deliberate thinking. Not all decisions benefit from speed. Some require reflection to avoid misinterpretation. Initial impressions should be revisited and discussed with those who produced the data. 
   

Mr. Unhyde: And none of this diminishes the value of data?

Dr. Concepto: Not at all. Data remains essential. But information does not interpret itself, and AI cannot replace human judgment. We arrive at understanding gradually, through reflection, dialogue, and an awareness of what we do not yet know.

Mr. Unhyde: So the challenge remains?

Dr. Concepto: Yes. The challenge is not necessarily to think faster or collect more data, but to think more clearly and methodologically with the support of analysts about what data mean. The quiet work of thinking—of investigating what something truly represents—remains one of the few human faculties that cannot be delegated, automated, or outsourced—ever.