Gartner published a press release today that puts a number on something most business leaders already suspect but haven’t been able to quantify: organisations with successful AI initiatives invest up to four times more in data and analytics foundations than their peers.

The press release, published April 16, 2026, draws on Gartner’s survey and analyst research across enterprise AI programs. The finding is not subtle. If you want AI to deliver returns, the investment in data quality, governance, and talent needs to come before — or alongside — the investment in AI tools.

The Gap Nobody Talks About

Here is the stat that should be on every CFO’s desk: 83% of CFOs report that less than half of their data, analytics, and AI investments have delivered financial results. And 80% of leaders say they struggle to track value from these investments at all.

These are not small numbers. They describe the majority of enterprise AI spending as underperforming with no clear way to measure how badly.

The disconnect is not a surprise if you understand how AI actually works. AI models, agents, and automation tools require clean, structured, well-governed data to produce reliable outputs. When that foundation is weak — fragmented systems, inconsistent formats, poor data quality, no governance — the AI on top of it compounds the problem rather than solving it. Garbage in, garbage out, now with a very expensive interface.

The organisations that are seeing returns have figured this out. They are spending on the infrastructure that makes AI work, not just on the AI itself.

What the High Performers Are Actually Doing

Gartner’s research from its 2026 Data and Analytics Summit provides additional texture on the spending patterns. Organisations with the highest satisfaction from AI initiatives invest roughly 1.78 times more in foundational capabilities — specifically data quality, governance, and analytics talent — than they invest in AI tools. High-ROI organisations spend four times more on process redesign and foundational change management than on the AI technology itself.

Read that again: the process and change management investment dwarfs the technology spend. The AI is not the expensive part of getting AI to work. The people, processes, and data infrastructure are.

This runs counter to how most AI spending conversations happen. Vendors lead with the technology. Software demos are compelling. New model announcements generate headlines. The boring, foundational work of building data pipelines, establishing governance policies, and upskilling teams does not get the same attention — but it is what separates the 83% that are not seeing returns from the organisations that are.

The Production Reality

Gartner’s research also found that just 8% of enterprises have agentic AI in production deployments. The gap between the boardroom conversations and actual deployed capability is significant.

It also found that 70% of agentic AI use cases will fail to deliver expected value, with the primary cause being wrong cost models at the outset. Businesses are calculating the ROI of AI based on tool costs and expected efficiency gains, while underestimating what it costs to build and maintain the data infrastructure those tools depend on.

This is not a technology problem. It is a planning problem. And it is solvable.

Why This Validates a Data-First Approach

Enterprise DNA has trained over 220,000 data professionals across 50 countries because we have believed for years that data literacy is the prerequisite for everything else — including AI. This Gartner study is the clearest external validation of that conviction we have seen.

The businesses that will capture the returns from AI are not necessarily the ones with the largest technology budgets. They are the ones that have invested in understanding their data, building the skills to work with it, and establishing governance that makes it trustworthy.

The specific skills that matter here are not abstract. Power BI, Python, SQL, data modelling — these are the capabilities that let your team build the data foundations Gartner is describing. AI tools can accelerate the work of someone who understands data. They cannot replace the foundation.

What This Means for Business

If you are a business leader evaluating your AI program — or wondering why the AI tools you have deployed have not delivered the returns you expected — the Gartner research points to a clear diagnostic question: what is the state of your data foundations?

Three things to check before spending more on AI tools:

Data quality. Can your teams trust the data coming out of your systems? If not, the AI will reflect that unreliability and amplify it at scale.

Analytics literacy. Does your team have the skills to understand AI outputs, build the pipelines that feed AI, and interpret what the models are actually doing? If your team is not data literate, you cannot evaluate whether your AI is working correctly.

Governance. Do you have policies around data access, quality standards, and accountability for AI outputs? Without governance, AI deployments become unmanageable as they scale.

None of these require a new platform or a vendor contract. They require investment in people and process — which is exactly what the Gartner research says high-performing organisations are prioritising.

If your team’s data skills need strengthening before your AI investments can deliver, Enterprise DNA’s learning platform is built for exactly this. Over 100 courses across Power BI, Python, SQL, and AI tools, designed for working professionals across all skill levels.

If you are a business leader who needs a clear picture of where your AI strategy stands relative to your data foundations, Omni Advisory is a practical starting point for that audit.