Who Owns Your Data When Everyone Does

Your CRM says you have 430 active customers. Your finance tool says you invoiced 380 last quarter. Your head of product pulled a number from the data warehouse last Tuesday and got something different from both. Nobody is lying. Nobody is wrong, exactly. The problem is that nobody owns the definition of "active customer," so every system applies its own.

This is not a tooling problem. You probably already have four tools that were each supposed to fix it.

The number that should embarrass the software industry

SAS surveyed 1,600 leaders of small and midsize businesses across 28 countries and found that 44.7% describe their data as scattered across different tools with no clear ownership or definitions. Only 11.8% report well-defined, owned, catalogued, and accessible data domains across their company.

Sit with that for a moment. Fifteen years of improving analytics platforms, better integrations, cheaper cloud storage, and the result is that fewer than one in eight SMBs has achieved basic domain clarity. If tooling were the answer, the numbers would look different by now. They don't.

SAS also identified data as the single largest structural barrier to scaling AI in smaller firms. Not budget. Not talent. Data. Specifically, data without ownership.

What scatter actually costs you

The cost of scattered data is not primarily the time your team spends arguing about reports, though that cost is real. The deeper cost is that every decision made from disputed data carries an invisible risk premium. You don't know which number is wrong. You don't know if any of them are right. So you either delay the decision until someone reconciles the sources, or you proceed and absorb the error downstream.

Siphambili and Nelufule, in their systematic review of data governance in smaller enterprises, conclude that governance is not optional for AI adoption in these firms. They also document why most governance attempts fail: existing structures are too complex, borrowed from large organisations, and collapse under the workload of small teams. The failure mode is not that founders ignore governance. It's that they import a design built for a 500-person data team and assign it to someone who also runs sales calls.

The case against naming an owner

The strongest objection to named data ownership is not that it doesn't work. It's that it creates a different problem.

In a team of eight people, the person you assign to own customer data also handles onboarding, sits in on product reviews, and takes the 9am call with your biggest account. Giving them formal responsibility for definitions and data quality does not give them additional hours. When they're overloaded or out sick, every data dispute waits for them, and the team works around the bottleneck in whatever way is fastest, which is usually the way that produces more inconsistency.

This failure mode is documented. Siphambili and Nelufule found that governance structures prove too complex for resource-constrained firms, and that complexity is precisely why uptake stays low. A governance title without matching capacity produces a role that exists on paper and nowhere else.

This objection is correct about the failure mode. It is wrong about the alternative.

The 44.7% of SMBs with scattered, unowned data have already chosen informal shared accountability. That is the baseline the counterargument is defending. The SAS data shows what that choice produces: data identified as the single largest structural barrier to growth, across 1,600 firms, in 28 countries. Informal coordination is not a lighter form of governance. It is the absence of governance, and the research treats it as the source of the problem, not a solution to it.

What named ownership actually requires

The version of domain ownership worth assigning is not a governance org chart. It is one person, one domain, one authority.

Pick your three most disputed data types. For most founder-led businesses, these are customer records, revenue figures, and product data. Assign one person to each. Give them two things: the authority to define what counts as a valid record in their domain, and the authority to reject data that doesn't meet that definition. Not a committee. Not a shared inbox. One person who is accountable when the number is wrong.

The OECD's cross-country analysis of SME data governance defines governance as arrangements that affect how data is created, collected, stored, used, and deleted. By that definition, informal shared accountability is not governance at all. It is the absence of arrangement. Naming an owner with real authority is the minimum move that creates an arrangement.

Siphambili and Nelufule call for a universal governance structure that reflects the realities of small enterprises. Their critique targets complexity, not ownership itself. A founder who assigns one person to own the definition of "active customer" and gives that person authority to enforce it has done something simple. It is not the large-enterprise governance design the review criticises.

The question you are actually avoiding

The reason most founders don't assign domain ownership is not that they've weighed the bottleneck risk and found it too high. It's that naming an owner forces a conversation about whose current definition is wrong.

If you assign your head of finance to own revenue figures, you are implicitly telling your head of sales that their pipeline report is not the authoritative source. That is a political conversation, and it is uncomfortable. Leaving ownership undefined lets everyone's numbers coexist without a verdict.

The SAS data suggests this tradeoff does not hold. Firms with undefined domains don't avoid the conflict. They defer it, repeatedly, into every meeting where someone pulls a number and someone else disputes it. The 11.8% of SMBs with well-defined, owned, catalogued, and accessible domains did not avoid the political conversation. They had it once, assigned the authority, and stopped having it at every quarterly review.

Start with customer data. Decide what "active" means. Name the person whose definition wins. Write it down somewhere the whole team sees it.