The critical data map: which datasets actually hold an operating company together
Management does not need all data - it needs a list of the information pillars of the business. How to find those pillars and what to do with them.
Almost every operating company has accumulated more data than it knows how to use. Yet the moment there is a serious disruption - a system outage, a key employee leaving, an urgent decision to be made - it turns out the needed data is not at hand, is out of date, or cannot be retrieved without hours of manual work.
The problem here is not volume. The problem is that nobody has asked the question: which data actually holds the operations together? Having a lot of data is not the same as knowing which data matters.
What critical data means
Critical data is not simply "important" or "confidential" data. It is data whose absence or inaccuracy directly stops or breaks an operational process.
A few examples that tend to appear on this list:
- Current price lists and terms with counterparties - without these, the commercial team cannot price deals correctly.
- Real-time inventory status - without this, production or logistics operates blind.
- The status of commitments to customers - which orders are in progress, what is overdue, what is awaiting confirmation.
- Financial limits and settlement status with key partners - without these, procurement or treasury makes decisions without a foundation.
- Reference and regulatory data that underpins calculations: rates, coefficients, classifiers.
Every company's list is its own. But the principle is the same: if this data is unavailable or wrong, the company starts operating on assumptions.
What critical data usually looks like in practice
At most operating companies I have encountered, the picture is similar. The critical data formally exists - it lives somewhere in the systems. But:
- it is updated manually, irregularly, and "the latest version is with Maria";
- it is duplicated in several places with discrepancies;
- getting it requires calling a specific person or running an export that takes several hours;
- when a system or responsible person changes, it gets lost or ends up living in Excel.
This is a normal stage of growth. But the longer it continues, the more expensive the absence of order in this data becomes.
Why this is a management question
Data about the state of operations is not a technical asset - it is a management tool. When a manager cannot get a current picture of key operational indicators without first asking several people, that is not an IT problem. It is a management problem.
I have seen situations where a major procurement decision was made on three-week-old data because there was no current version. Or where the commercial director did not know the actual state of receivables because the system export ran once a month.
These situations are not accidents. They are the result of nobody having identified which data is critical and ensured its availability.
How to build a critical data map
The practical approach here is not complicated. It has a few steps.
The first step: ask the leaders of key functions - what data do you need to make operational decisions every day? Not "it would be nice to have", but specifically "what leaves you working blind without it".
The second step: for each such dataset, determine where it lives today, how current it is, who is responsible for it, and how quickly it can be retrieved.
The third step: prioritize - which of these datasets are critical (failure equals operational risk), and which are merely convenient.
The fourth step: for the critical datasets, define the requirements - what freshness is needed, who is the source of truth, who is accountable for quality.
This is not a large project. It is a conversation with people and a diagram on paper. But it is exactly what becomes the foundation for subsequent data work.
Three questions as a quick test
To quickly assess the situation in a company, ask three questions:
- If I need to know the actual state of a key operational indicator right now - how much time and whose involvement does that take?
- Is there data in the company whose accuracy formally belongs to nobody? Who owns the number in the report is a question that surfaces every time this kind of data is missing.
- If the person who "knows where everything lives" leaves - what exactly does the company lose?
The answers to these questions usually show whether the company has a managed data map, or just a set of habits.