Data ownership: who signs off on the number
In most companies data exists but no one is responsible for its quality. I look at what data ownership actually means in practice.
The same situation comes up in meetings repeatedly. Someone quotes a number - a customer count, a conversion rate, a sales figure. Someone else says: "I have different data." Then comes the effort to establish whose number is correct. It turns out everyone has their own export from their own system with their own methodology. The meeting moves on without a resolution.
This is not a data problem. It is an absence of data ownership.
What "owning data" means
Data ownership is not a technical term. It is an organisational role.
A data owner is a specific person who is accountable for the data in their area being correct, current, and available to those who need it. This is not the person who stores the data. It is the person who signs off on what it says.
The owner of customer data is accountable for the definition of "active customer" being clear, documented, and applied consistently everywhere. The owner of financial data is accountable for revenue being calculated one way, not five different ways.
Why this rarely exists
Companies invest money in data tools - warehouses, BI systems, analytics. They rarely invest in the organisational side.
Several reasons. The first is that this is not an obvious role. It is unclear where it lives in the org chart: in IT, in finance, in a specific business unit? The second is that it is work without a visible result. Good data ownership looks like the absence of problems, not like an achievement. The third is conflict of interest. If a unit is responsible for data, it is also responsible for acknowledging errors in that data.
The result: data exists, systems exist, but there is no one accountable for quality.
What happens without a clear owner
Several typical consequences:
- different units calculate the same metric differently and do not notice the discrepancy until it surfaces in a visible way;
- when a new system or integration is introduced, it is unclear who is the source of truth - where to get data from and who is accountable for its correctness;
- data errors are discovered by accident and late - after they have already affected decisions;
- there is no process that responds to degradation in data quality - nobody knows what to do when data gets worse.
How to set this up
You do not need to create a separate department or launch a large programme. You can start small.
First step - make a list of the key business data sets. Not all data in the company, but the data used to make decisions: customers, sales, finance, operational metrics.
Second step - for each type of data, name a specific owner. Not "the finance department in general", but a specific person by name.
Third step - for each type of data, document: what is included in this concept, where the data comes from, who has the right to modify it, how often it is updated.
This is not a project. It is a conversation that can happen in a few meetings. But the result - a clear answer to "whose number is this?" - changes the quality of working with data in the company.
Data without an owner is no one's data. And no one's data is useful to no one.