The Missing Data Link: How Operating Models Hold Back Data Value
Most organizations are investing heavily in data, yet very few manage to scale their data products in a way that creates real enterprise value. McKinsey captures the issue with unusual clarity: “Building valuable data products is much less of a technical challenge than a strategic and operational one.” The bottleneck is not technology. It is fragmentation, governance, and the absence of a scalable operating model. This tension, between ambition and execution, defines the entire problem.
Data Products Are Built… but Not Scaled
Across industries, companies continue to operate in a single-engine, single-car logic. Each use case receives its own pipeline, its own data preparation, its own governance. The result is predictable: duplicated work, inconsistent data, slow delivery, rising costs, and a growing sense of frustration among business teams. McKinsey notes that this approach leads to “fragmenting data programs that fail to scale or generate the value expected.” The paradox is striking. The more organisations invest in data, the less value they seem able to extract from it.
The Economics of Data Are Broken
McKinsey introduces a concept that deserves far more attention: the data flywheel. A data product becomes economically powerful only when it is reused across multiple use cases. Reuse lowers marginal cost and accelerates time-to-value. Once a data product supports five use cases, costs fall by roughly 30 to 40 per cent compared with building separate pipelines. Yet most organizations never reach this point. They build too many data products. They fail to cluster use cases. They misalign incentives. They underinvest in DataOps. They treat data as a project rather than a product with a lifecycle. Without reuse, the economics collapse. The flywheel never turns.
The Real Bottleneck: The Operating Model
McKinsey is explicit: the obstacle is organisational. Ownership is unclear. Data Product Owners lack authority. Business teams arrive too late in the process. No marketplace exists for consumption. Governance does not encourage reuse. “A data product program viewed as ‘just an IT project’ won’t succeed in creating value.” The opportunity, however, is substantial. Organisations that redesign their operating model unlock faster delivery, lower costs, and higher adoption, the three pillars of scalable data value.
The COO Lens
This is where the conversation shifts. The question is no longer how to build better data products. The real question is how to redesign the operating model so that data products can scale. The COO Lens reframes the challenge. It asks whether use cases are clustered to maximise reuse. Whether a lifecycle owner is accountable for value. Whether friction is being removed from data flows rather than added. Whether a DataOps backbone exists to automate the majority of the work. Whether incentives reward reuse rather than reinvention. None of this is a data problem. It is a problem of process, governance, and operating design, the core of Data & Process Redesign.
Conclusion: The CEO Perspective
For CEOs, the message is simple. Data products do not scale by accident. They scale by design. The organisations that succeed will be those that treat data as a product, build for reuse, invest in DataOps, empower product owners, and redesign processes around value rather than silos. The winners will not be the companies that build the most data products, but those that create the conditions for data to compound.
To explore how organisations can redesign their operating model to unlock scalable data value, the full framework is available here:
Data & Process Redesign → https://christopheschmid.com/data-process-redesign.html