This article is adapted from the first chapter of Solved on Paper but Stuck in Practice, a new research report published by Corinium and Quest. You can read the full report here.
___
Enterprises across major industry verticals have been investing in their data architecture for many years, and almost all now have the necessary tools in place.
97% of organisations now use a formal data modelling tool according to our research, with half reporting deployment across the enterprise and the rest constrained within specific teams.
However, adoption alone does not guarantee consolidated or uniformly enforced processes. Only 43% of respondents to our research create and maintain their models with fully standardised tooling and no manual overlay.
What’s more, 44% still rely on a mix of tools and manual processes to get the job done.
According to Glenda O’Keefe, Field CTO at Quest Software, the tooling question itself is largely solved, but that has shifted the real challenge elsewhere:
“Adoption is no longer a training problem; modern modelling tools are intuitive and AI-assisted. It’s a migration problem—having to move years of legacy models’ definitions and pipelines into a unified platform,” she says. “Migration competes with new delivery work for budget and time. That’s why so many organisations remain stuck in the middle of the maturity curve.”
A similar pattern emerges in data governance at scale. Only 9% of respondents describe their governance as centralised and enforced, and 70% only report governance “defined in some areas.”
The same is true of the standardisation of data modelling where only 25% of respondents report standardised practices adopted across the enterprise, while 58% have them adopted only in some teams.
Buying the tool did not, on its own, buy consolidation.
O’Keefe places most organisations somewhere in the middle:
“They’ve purchased the tools, they’ve drafted the standards, they’ve even begun pockets of governance, but they haven’t unified them. Until modelling becomes a shared business capability rather than a technical task, consolidation will likely remain incomplete and progress will stay uneven.”
Elizabeth Press, Founder of D3M Labs and a Deputy Chief Digital Officer, points to a more human explanation for the same fragmentation:
“A lot of tool use is tied to individual managers. A manager brings in a vendor, does the implementation. That’s their project. Then that manager leaves, and the next one comes in with their own ideas,” she says.
Solving tool redundancy requires an enterprise view, but is hard to do for all possible use cases, according to Sally Bashuan, SVP, Global Head of Data Governance at Federated Hermes:
“You need someone to step back and take an enterprise view of the data, someone focused on solving the business problem,” she says. “But I don’t think you’ll get rid of it altogether because I think there are unique business cases.”
Our research also hints at a lack of collaboration on the human side. Only 31% of organisations say business and technical teams collaborate on data models ‘to a great extent’, and 2% report no collaboration at all.
Pier Martin, Vice President, Data and Analytics at ZEAL Network SE, sees this gap appear in how teams talk about each other: “Warning signs often appear in language. When teams describe each other as obstacles rather than partners, that’s a sign the relationship itself needs attention.”
The data introduces an intriguing pattern crossing data modelling, governance and the data stack. Many organisations have the necessary tools in place. But most have not deployed them consistently across the organisation, and very few have achieved deployment at enterprise scale.
_
This article is an excerpt from Solved on Paper but Not in Practice. Download the full report to explore why data modelling maturity often stalls before it scales, and how organisations can close the gap between the tools and standards they have in place and the consistent practices needed to deliver usable data for analytics and AI.