Clean Does Not Mean Perfect
The phrase “data cleaning” can make people quietly despair. It sounds like a vast technical project, the sort that begins with optimism and ends in a spreadsheet called Final_Final_v7. HR teams are already busy. Operations teams are already stretched. Who has time to make every record perfect?
The good news is that perfection is not the aim. Clean HR data means information good enough to support sensible decisions. It means managers can trust the start date, payroll can trust the salary, compliance can trust the training record, and leaders can trust the headcount report. It means the business is not making decisions from guesswork dressed as data.
The 45-minute inventory is designed to create movement. It will not solve everything. It will show where the problems are.
Three Versions of the Truth
Most organisations have data clones: different records that claim to describe the same reality. The payroll spreadsheet says an employee started on 3 March. The HR system says 1 March. The training log says they joined in April because that is when someone first added them. Which date is true? It depends whom you ask.
Now imagine the same problem with job titles, working hours, certifications, emergency contacts, probation dates, visa details or equipment lists. Suddenly a small inconsistency becomes an operational risk.
The problem is rarely that people do not care. More often, each spreadsheet was created for a reasonable purpose. Payroll needed one view. Operations needed another. A manager built a local tracker because the official system was slow. Training kept its own log because nobody trusted the HR export. Each workaround made sense at the time. Together, they created confusion.
Why Dirty Data Damages Trust
Bad HR data is not an abstract issue. It affects people directly. An employee paid late does not experience a “data quality concern”. They experience anxiety. A manager missing a probation date does not experience an “information gap”. They experience embarrassment and lost credibility. A training certificate recorded incorrectly may affect compliance, client trust or even safety.
It also damages the company’s identity. If the organisation says it values care but cannot keep basic employee information straight, people notice. If it values clarity but managers rely on hidden spreadsheets, people adapt by creating more hidden spreadsheets. If it values trust but reports are always questioned, decision-making slows down.
Data quality is culture in another form. It shows whether the company respects the facts on which people depend.
The 45-Minute Inventory
Set a timer. Invite HR, Operations, payroll and one or two managers who understand the daily reality. The aim is not to debate the perfect future system. The aim is to list where people data lives now.
Start with broad categories: employee master data, contracts, payroll details, training records, absence records, performance notes, equipment records, access permissions, emergency contacts, onboarding checklists and offboarding records.
For each category, ask six questions. Where is it stored? Who owns it? Who uses it? How reliable is it? Is it duplicated? What should happen next?
The answers will not be tidy. That is fine. The mess already exists. The inventory simply stops it hiding.
Choosing a Source of Truth
Every important data category needs a trusted home. That does not mean every team loses access to the information they need. It means the business agrees which record is authoritative.
For example, the HR system may be the source of truth for employee status and contractual details. Payroll may be the source of truth for payment information. The learning system may be the source of truth for completed training. A manager’s spreadsheet may be useful for planning, but not for official employee records.
This distinction matters. A spreadsheet can be a working view without becoming a second truth. Problems begin when a working copy becomes an unofficial authority.
Retiring What No Longer Helps
Some spreadsheets should be improved. Some should be integrated. Some should be retired. The question is not whether a spreadsheet is bad. A well-designed spreadsheet can be perfectly useful. The question is whether it is trusted, owned, current and necessary.
If nobody owns it, it will decay. If it duplicates a better source, it will confuse. If people use it because the official system is painful, that is a signal to improve the system. If it exists because one manager likes a private view, it should not become a business dependency.
The Human Side of Cleaning
Data cleaning can feel accusatory. People may worry that their local tracker will be criticised, or that mistakes will be blamed on them. It helps to frame the exercise carefully: “We are not looking for fault. We are looking for friction.”
The best question is not, “Who created this mess?” It is, “What pressure made this workaround necessary?”
That question reveals the system. It may show that people need faster reports, clearer forms, easier access, better training or a more sensible approval path. Clean data is not achieved by scolding staff. It is achieved by designing a process people can actually follow.
Connecting Data to Purpose
What does a data inventory have to do with mission and belonging? More than it first appears. People feel respected when the organisation knows who they are, what role they do, what they need, what they have completed and what support is due to them. They feel less respected when they must repeatedly correct basic information.
A company’s vision may be ambitious, but ambition relies on operational truth. How can leaders plan growth if headcount data is unreliable? How can managers develop people if training records are incomplete? How can employees trust the company’s care if their own details are mishandled?
The 45-minute inventory is not glamorous. It is one of the most practical ways to rebuild confidence.
A Practical Sorting Method
Once the inventory is complete, sort each data source into one of four categories. Keep, improve, merge or retire. Keep the records that are trusted, owned and useful. Improve the ones that matter but contain gaps. Merge duplicates where two teams are maintaining similar information. Retire files that nobody trusts, nobody owns or nobody should be using.
This is where leadership support helps. People can become attached to their trackers because those trackers have protected them from poor systems. Do not simply take them away. Ask what need they meet. Then make sure the source of truth meets that need better.
It is also wise to avoid a grand clean-up that never ends. Choose one high-risk area first. Payroll details, training compliance or onboarding records are often good candidates. Clean enough to improve decisions, then move to the next area.
The Identity Test for Data
Data can seem impersonal, but employee data is personal by nature. It describes someone’s role, pay, development, working pattern, absence, performance and history with the company. Handling it well is part of respecting them.
Ask: would an employee feel confident if they saw how their data is managed? Would they see care, clarity and ownership? Or would they see scattered files and quiet uncertainty?
If the company wants employees to feel part of something purposeful, the organisation must treat their information as part of that relationship. People are not entries in a spreadsheet. But if the spreadsheet is wrong, they are the ones who feel it.
