The Death and Rebirth of Data — Part 4

In this fourth chapter of The Death and Rebirth of Data, we shift from the question of who owns data to an even more fundamental question: What does using data actually feel like?

Over the last decade, organizations have improved pipelines, platforms, and products, but the experience of interacting with data has remained painfully unchanged.

This blog explores why the future of data is not just technical but experiential, and why real value emerges only when data becomes intuitive, conversational, and human.

In Part 3, I explored how data slowly drifted out of the hands of the business, becoming a technical asset owned by technologists rather than a strategic asset owned by decision-makers. Restoring that ownership is essential.

Gartner has repeatedly highlighted this drift in its Data & Analytics Leadership Vision reports, noting that business ownership is the #1 predictor of data initiative success.

But ownership alone isn't enough. Even if the business owns the data, and even if engineering builds the pipelines, the question remains, what does interacting with data actually feel like?

BARC Germany's Data & Analytics Trend Monitor notes that "ownership without usability" is one of the top reasons business stakeholders disengage from data programs.

And this is where the industry has fallen painfully short. Over the last few years, the idea of "data as a product" has gained traction. It's a powerful concept. Treat datasets, metrics, and analytical outputs like physical products, with the same level of attention, care, and strategic importance, with clear owners, SLAs, documentation, and customers.

This framing was introduced and formalized by Zhamak Dehghani at ThoughtWorks, where the modern Data Mesh principles first emerged.

But as valuable as that shift is, it only solves half the problem. A product is not enough. You can build a beautiful, well-documented data product, but if the experience of discovering it, understanding it, or using it is a nightmare… it won't matter.

McKinsey's 2024 "State of Data Transformation" found that 63% of business users abandon data tools due to poor usability, not poor data quality.

Data suffers from this same problem. We've created products, not experiences. We've focused on metadata, not meaning. We've built systems, not sensations.

MIT Technology Review and Databricks jointly noted in 2025 that "the analytics industry has optimized for storage and compute, not user experience."

Data needs to be experienced, not consumed. Most data teams assume that "using data" is a rational, information-consumption activity. But in reality, data is deeply experiential. That is why the question-and-answer approach of generative AI is accepted so naturally. Humans are taught to learn that way. Ask a question, get answer, explore with another question.

This aligns with findings in Stanford's 2024 Human-AI Interaction research, showing that conversational interfaces match innate cognitive learning patterns.

When we look closely at why so many data initiatives quietly stall, the cause is rarely the data itself — it is the friction surrounding it. Every interaction with a dashboard, every search for a metric, every clarification ticket is a micro-moment of experience, and those moments compound. What looks like a "data quality" problem is often, in truth, an experience-design problem. The Friction Chain below traces how that erosion happens in sequence: a slow interface drains patience, ambiguous definitions drain trust, a frustrating journey drains belief in the answer, and rigid tools drain the curiosity that made someone open the tool in the first place.

Each link in that chain is small on its own, but together they decide whether data becomes a habit or a hassle. If the interface is slow, people give up. If the definitions are confusing, people lose trust. If the journey is frustrating, people revert to gut decisions. And if the tools feel rigid, people stop being curious altogether — which is the most expensive failure of all, because curiosity is the engine of every insight a business will ever generate. The four cards below make those failure modes explicit, not as edge cases, but as the default experience most organizations are unintentionally shipping to their own people every day.

Forrester's CX Index identified "slow insights" as one of the top barriers to data-driven decision-making, causing reliance on intuition over evidence.

Data is not just something people receive. It's something they experience through every click, search, question, and the decision they are making.

And that experience has been largely neglected.

Unfortunately, the traditional data experience is broken by design. The standard experience today within a majority companies would entail opening a BI tool, navigating dashboards someone else built months ago, hope metric definitions still match reality, realizing their question isn't answered anywhere, logging a ticket, waiting, getting a partial answer, asking for a revision, wait some more, and giving up and exporting to Excel.

This pattern is documented in Gartner's Magic Quadrant reports for BI tools: dashboards are outdated by an average of 30–90 days at most enterprises.

We've made data feel like filing tax returns, slow, bureaucratic, confusing. It punishes curiosity instead of encouraging it.

If data is to be an experience, we need a different mindset. We need data that behaves like an intuitive digital product.

What organizations are beginning to realize is that the problem was never simply access to data, it was the experience surrounding that access. Traditional BI environments were built around technical delivery models, not around the speed, context, and flow of modern business decision-making. The user was expected to adapt to the platform instead of the platform adapting to the user.

Experience-first data products reverse that model entirely: instead of forcing business users through layers of dashboards, tickets, governance documentation, and technical interpretation, the interaction becomes immediate, contextual, and intuitive. Data begins to behave less like a disconnected reporting system and more like a trusted digital product designed around the human decision-making process itself. This is where the shift from delivery to decision becomes real. Where governance, context, AI, usability, and business intent converge into a single experience that feels natural rather than operationally exhausting.

The shift from dashboards to conversational data is highlighted in McKinsey's 2024 "The AI-Centric Enterprise" report, which argues that AI will become the primary UX for data.

A true semantic layer is not only a technical construct. It becomes an experience layer where business meaning is unified, accessible, and human-friendly. Built around how the business thinks, decides and acts, not around how data happens to be stored.

When the experience layer is designed first, the semantic layer stops being a back-office translation table and starts behaving like a living interface between the organization and its data. Definitions are shared, context travels with every metric, and the same question asked in Slack, in a CRM, or in a boardroom returns the same trustworthy answer. This is what turns a stack of pipelines, warehouses and models into something a business can actually use to move from delivery to decision: meaning becomes navigable, trusted, and operational rather than negotiated through layers of dashboards, tickets, and translation.

In this model, the semantic layer no longer exists solely to standardize reporting definitions. It becomes the operational bridge between governed data products, AI interaction, and human decision-making. The experience layer allows business meaning, governance, context, and usability to travel together, enabling trusted decisions without exposing users to the underlying technical complexity.

This aligns with the modern semantic layer perspective outlined by dbt Labs, AtScale, and the Gartner Market Guide for Semantic Layers.

It lets anyone ask "Show me churn risk for my top customers" without knowing which table or field to query, or which AI model to use.

Data as an experience matters. No matter how modern your pipelines, warehouses, lakehouses, or dashboards, if the business cannot experience data effortlessly, you have not created value. Data becomes valuable when someone feels confidence in a decision, clarity in a trend, insight in a pattern, speed in getting an answer, or trust in a metric.

Harvard Business Review emphasizes in "The Cognitive Era of Decision-Making" that value is created only when insights translate into emotional confidence and business action.

What organizations are ultimately searching for is not another dashboard, pipeline, governance document, or AI model. They are searching for confidence. Confidence that the data is trusted. Confidence that the answer is relevant. Confidence that the context has not been lost between systems, teams, and time. This is why the future of data is no longer just architectural, it is experiential. The organizations that succeed will be the ones that design data products around how humans actually think, decide, collaborate, and act. In that world, governance becomes visible, semantics become human, AI becomes assistive, and trusted data becomes part of the natural rhythm of business rather than a disconnected technical exercise.

• BARC Germany (2025). Data & Analytics Trend Monitor. BARC Institute.
• Databricks & MIT Technology Review (2025). The Data Paradox: Why Modern Analytics Still Fails Users.
• Dehghani, Z. (2021). Data Mesh: Delivering Data-Driven Value at Scale. O'Reilly Media.
• dbt Labs (2024). The Modern Semantic Layer: A Market Perspective.
• Forrester Research (2025). Customer Experience Index: Data-to-Insight Bottlenecks.
• Gartner (2024). Data & Analytics Leadership Vision.
• Gartner (2025). Magic Quadrant for Analytics & BI Platforms.
• Harvard Business Review (2024). The Cognitive Era of Decision-Making.
• McKinsey & Company (2024). The AI-Centric Enterprise: Seven Shifts Defining the Future of Work.
• O'Reilly Media (2023–2024). Semantic Layers and the Evolution of Analytical Experience.
• Stanford University (2024). Human-AI Interaction in Decision Systems.