Data products people use were the goal. AI needs products it can trust at the point of decision.

The Signal

Hazel Saunders, Business Editor at The Ritz Herald, recently published How Enterprises Can Design Data Products That Business Teams Actually Use. Her argument is a good one. A data product can pass every technical review and still deserve to be retired if it has no dependable place in anyone's work. Success is not technical completion. It is whether the product is designed around how business teams actually work, with a named user, a real job to be done, genuine ownership, usable access and a clear path toward workflow dependency.

Saunders is right to move the conversation away from whether a data product technically exists and toward whether it becomes part of how the business works. Our extension is that AI makes that question considerably more consequential.

The question is no longer simply whether people will use a data product. It increasingly becomes whether people and AI can depend on it when a decision needs to be made.

Data products people use were the goal. AI needs products it can trust at the point of decision.

That is a higher standard, and it is why the data product itself has evolved. It now has to carry not just the relevant data, but enough approved business understanding for a person or a machine to act on it responsibly.

Adoption Was Never the Ceiling

Organizations have traditionally measured data products by discovery, adoption and reuse. Saunders pushes further, to workflow dependency, where a recurring business process relies on the product and removing it would create visible friction. AI introduces one more threshold beyond that.

Discovery → Repeat Use → Reuse → Workflow Dependency → Decision Dependency

At decision dependency, people and machines are no longer simply consuming a product. They are depending on it to inform or support a decision. The practical test is to imagine switching a product off tomorrow. What decision becomes harder? What manual reconciliation returns? Which AI agent loses a trusted input? If the honest answer is essentially nothing, usage figures are telling you very little about value.

Gartner arrives at a related conclusion from a measurement perspective. Its May 2026 research argues that technical indicators alone do not demonstrate the business impact of data products, and recommends outcome-driven metrics tied to realized value. Our reading is that the outcome worth measuring is the decision itself.

Business-Led Data Products

Getting to decision dependency usually runs into the delivery model. A business team identifies a need. Requirements are written and interpreted. A technical team interprets them again. Something is built, returned for testing, and the business works out whether its original meaning survived the translation. That cycle can produce excellent products when the organization has time. AI compresses the time available, and each translation step quietly removes business nuance. An experienced supply chain manager knows which supplier delay is ordinary and which is a genuine warning. A finance leader knows why two nearly identical measures are approved for very different uses. Much of that never lives cleanly in a database.

This is why business-led data products matter to us. Business-led does not mean technology-free, and it does not mean bypassing data teams, eliminating engineering or relaxing governance. Technology teams still provide and operate the trusted enterprise foundation. Governance teams still set policy and controls. What changes is that the people closest to the decision can participate directly in defining what the product means, which data matters, what context is important and which policies apply, rather than continually translating their knowledge through requirements, tickets and code.

Business knowledge should not have to be translated into code before it can become enterprise capability.

A finance expert should not need to become an engineer before finance knowledge becomes reusable, and a customer leader should not need to become an engineer before customer understanding becomes consumable by AI. The Decision-Driven Enterprise describes this as the shift from engineer-led delivery, where the business raises a ticket and waits in a queue, to business-led delivery, where the people who own the decision can build and own the data product.

Trusted Business Understanding

Data access and business understanding are not the same thing. A model can be given access to a hundred tables and still not know which definition of revenue applies to the decision in front of it, which customer classification is authoritative, whether policy permits a sensitive field to be used in a particular action, or which supplier exceptions actually matter operationally.

Enterprise AI depends on what a measure means, which definition applies, who owns it, how good and how current it is, where it came from, which policies restrict its use and whether it is suitable for the decision being made. That is what we mean at Data Tiles by trusted business understanding, and it is the difference between data the organization stores and knowledge the organization can safely act on.

AI should inherit the understanding instead of being expected to infer it.

As Cameron Price argues in Building Intelligent Organizations, most enterprises can now build AI faster than they can explain themselves to it. Models and agents change. The accumulated business understanding of the enterprise is far more specific and far harder to reproduce, which makes it the asset worth capturing once and reusing across people, applications, analytics and AI.

Active Governance

Governance cannot simply document what should happen upstream. If a data product is going to support people and AI at the point of decision, then context, policy, ownership, quality, lineage and permissions need to remain connected to the product as it is used, whether it is consumed in a report, an API call, an application or an agent. That is the purpose of active governance.

The distinction matters more with AI because of how the risk changes. An AI answer can look convincing and still be wrong. A person who encounters ambiguity tends to stop, call someone or question a definition. An agent can act repeatedly at machine speed. Once AI moves from answering questions to participating in workflows and decisions, the issue is not whether it can reach data. It is whether it has the right data, with enough business understanding and governance to use it appropriately, and whether the resulting decision is traceable and defensible.

What the Market Evidence Shows

The market is moving into broader data product adoption while still struggling to prove commensurate value. BARC and Actian report that operational use of data products increased from 48% in 2024 to 69% in 2026, and that organizations with company-wide data products were 3.4 times more likely to successfully move AI projects into production. KPMG's survey of 250 executives found that only 35% reported achieving extensive value from their data product initiatives. Enterprises are becoming better at having data products than at proving what those products changed.

Correlation is not proof of causation, but the direction makes sense to us. Organizations that already know how to package data with ownership, quality expectations and reusable structure are better positioned for AI than organizations rebuilding trusted inputs for every new project.

The economics point the same way. In The Missing Data Link, McKinsey describes a flywheel in which reuse lowers incremental cost and accelerates value capture. At one telecommunications company it estimated that 60% to 80% of the initial effort spent finding, preparing and quality-assuring data was one-time work. At an international consumer company, a reused data product supporting five use cases had projected costs approximately 30% lower than building five individual pipelines. The objective is not maximum product count. It is maximum trusted reuse, which matters more once every new agent and use case would otherwise rediscover the same definitions, policies and ownership from scratch.

Start With the Decision

All of this leads back to a question we think executives should ask before any of the technical work begins: what decision are we trying to improve? From there the organization works backward to the trusted data, business context, semantics, governance and evidence required to support that decision.

Saunders makes a version of this point well. “Sales analytics” is vague. Giving regional sales leaders a trusted view of renewal risk before a Monday pipeline review is specific enough to design against, because it identifies the user, the timing, the purpose and the consequence of failure. That is the difference between being data-driven and becoming a Decision-Driven Enterprise, and it moves the purpose of data from supply to decision.

What We See Across APJ

In conversations with customers and partners across Asia Pacific and Japan, the pattern is consistent. There is rarely a shortage of data, and rarely a shortage of platforms. Organizations have already invested substantially in cloud, data platforms, catalogs, governance, analytics, engineering and increasingly AI infrastructure. The investment already exists. Yet the person responsible for a decision can still be waiting.

The infrastructure is not necessarily failing. The gap sits between that infrastructure and the moment somebody actually needs to decide. The opportunity for most organizations is getting more business value from what they have already built by connecting trusted data and business understanding more directly to decisions.

How Data Tiles Operationalizes This

Latttice is our zero-code Data Product Workbench. It allows business and domain teams to work alongside data and governance teams to assemble governed, reusable data products across the existing enterprise environment, carrying the context required for real business use. Context is captured while it is still in the hands of the people who understand it, and active governance travels with the product.

Lenz extends that trusted foundation into AI, enabling agents to operate from governed data products and the business context attached to them rather than inferring meaning from whatever data they can reach.

Latttice establishes trusted business understanding. Lenz enables AI to work from it.

The Market Signal

Hazel Saunders asks whether business teams actually use the data products organizations create. It remains exactly the right question. Our extension is that AI adds another one. Can people and AI depend on those products when a decision needs to be made?

The next generation of successful data products will not be judged simply by whether they are discoverable, usable, reusable or governed. They will be judged by whether they carry enough trusted business understanding to become dependable at the point of decision.