Your data strategy is not a technology problem

The conversation about data strategy in most organisations begins with tooling. Which data platform? Which catalogue? Which governance tool? Which visualisation layer? These are not the wrong questions. They are the wrong first questions. Asked before the organisational questions, they produce a technology investment that does not solve the problem it was procured to address.

The organisational questions that have to come first

Data quality problems are almost always ownership problems before they are technical problems. The data is duplicated because two systems both claim to be the source of truth for customer records and nobody has made the call about which one wins. The data is incomplete because the team that generates it has no downstream visibility of what happens when it is wrong. The data is inconsistent because there is no shared definition of what a customer, a product, or a transaction means across the organisation.

A data platform does not resolve ownership disputes. A data catalogue does not create accountability for data quality. These require decisions about who owns what, what the canonical definitions are, and what happens when those standards are not met.

Data ownership is not data stewardship

Most organisations have data stewards. They attend governance meetings and document data definitions. They rarely have authority to change how data is generated, stored, or used. Data ownership requires authority, not just accountability. The data owner for customer records needs to be able to mandate changes to the systems that generate those records, not just document the problem.

The domain model has to precede the technology model

Before designing a data architecture, the organisation needs a shared understanding of its business domain model: what are the core entities, what are the authoritative sources, what are the agreed definitions. This is a business exercise that requires business decision-making, not a technology exercise.

Start with the decisions, not the infrastructure

The most effective way to begin a data strategy is to identify the three to five decisions the organisation most needs to make better, and trace backwards to the data those decisions require. The gaps that surface in that exercise define the priority data investments. They may or may not require new infrastructure. They will almost certainly require organisational change.