The winners in the AI era won’t necessarily be the organisations with the best AI models. They’ll be the organisations with the most trusted, accessible, governed and business-ready data estates.
That’s the view of Gunther Wucherpfennig, Executive: Data Analytics and AI at Mint Group, who notes that after years of AI hype, organisations are discovering that the real determinant of success is not the sophistication of the model, but the maturity of the data estate beneath it.
Wucherpfenning uses a simple analogy to explain why so many AI projects underdeliver. Imagine buying a bestselling business book: a compelling title, a respected author, a dynamic cover. A hundred copies are handed out to the sales, marketing and leadership teams. The first few chapters read well – but by chapter three, the content starts to unravel: contradictory sections, misspelt names, numbers that don’t match from page to page.
The team stops trusting the book. Not because the book itself was poorly conceived, but because the content turned out to be unreliable.
“That’s exactly what happens when organisations deploy AI on top of poor-quality data,” says Wucherpfennig. “The model can be world-class, but if the underlying data is incomplete, duplicated, outdated or inconsistent, the AI has nothing trustworthy to work with.”
The result, he says, is inaccurate outputs, contradictory recommendations, poor decisions and users who quietly stop trusting the system, resulting in stalled adoption and the evaporation of business value.
Global research firms regularly report low success rates for AI implementation – around 40% according to Wucherpfennig. And the root cause is almost always the same: data quality, data governance and data accessibility.
What 'AI-ready' actually means
Rather than starting with the AI model, Wucherpfennig argues that the message for technology leaders is straightforward: AI readiness is fundamentally data readiness.
Organisations should therefore start with an honest AI-readiness assessment across six areas:
- Data quality – is the information accurate, current and consistent?
- Data integration – can your systems talk to each other, or are you sitting on disconnected silos?
- Security – is protection built into the data platform by design, using principles like zero trust, role-based access and multi-factor authentication, rather than bolted on afterwards?
- Governance – who owns the data, who can see it, who can change it and how is that tracked and audited?
- Metadata and discoverability – AI cannot use data it cannot find.
- Infrastructure scalability – the ability of your IT infrastructure to grow or shrink as business demand changes, without having to redesign or replace everything.
To this list he adds an often-overlooked area and that is business readiness – the processes, operating model and culture needed to adopt AI responsibly.
“Technology alone is insufficient,” he says.
Bridging ambition and value
This is where a modern, unified data platform earns its keep. Generically, the capability organisations should be looking for combines several things in one environment: bringing data from disparate systems – finance, CRM, point-of-sale, legacy platforms into a single, normalised foundation; built-in governance and security rather than governance as an afterthought; reduced duplication, so the same customer or product record isn’t scattered across 17 versions of the truth; and native support for the analytics and AI tooling built on top of it, from real-time reporting to AI agents.
Done well, this shifts data teams away from manual, repetitive integration and cleansing work – historically a multi-year, many-person effort – towards higher value work: managing, refining and extending what the platform automates, allowing teams to focus on value creation, not plumbing.
The real message for technology leaders
AI is not the silver bullet. It is the engine. The fuel is the data. And without the right fuel, even the most advanced AI initiative will stall.
Wucherpfennig’s recommendation is clear: organisations must invest first in data readiness, not AI experimentation. Those that do will unlock the real promise of AI – trusted insights, faster decisions, stronger governance and competitive advantage: measurable business value.
Those that don’t will continue to buy the “best book in the world” only to discover that nobody trusts what’s written inside.
Share
