A missing manager, a duplicate email, or an exit date that comes before the joining date does not stay a small mistake. Approvals stall. Attendance is marked wrong. Leave cannot be applied. Headcount looks fine when it is not. Data Health shows those gaps while you can still fix them.
It reads the people data you already keep. It does not change anything by itself. You open the list, go to the person or department, and fix it where you normally would.
Healthy, needs attention, or at risk
Names, teams, and what is wrong
A link to the screen where the record lives
Download when you need a working file
Same check. Three reasons it comes up when a company is deciding whether its people data is safe to trust.
Most firefighting starts with a record that was never finished. Someone has no manager. Their shift was never set. Two people share a work email. Probation ended last month and nobody confirmed it. Data Health groups those problems, shows how many people are affected, and lets you open the list. You fix the record in the employee or organisation screen you already use. Refresh, and the count should fall.
If someone has left but still shows as active, or their exit date is before they joined, every people number in the room is softer than it looks. Leadership does not need a technical audit. They need a plain label - healthy, needs attention, or at risk - and the confidence that HR can name the records behind it. This sits on Professional and Enterprise because it is an operating control, not a starter extra.
Managers feel this when leave or attendance corrections never arrive. The cause is often upstream: no manager, a manager who has already left, or a reporting loop that sends the request in circles. Data Health is an HR screen, not a manager inbox. What managers get is fewer “why is this stuck?” conversations once HR has cleaned the line.
Companies already store joining dates, managers, shifts, and contact details. What fails is noticing the holes before they stop a request, a leave application, or a headcount conversation.
Wait until an employee complains
By then the approval is late, the leave is blocked, or payroll week has started
A refreshable list of the gaps, with the people named
A spreadsheet audit once a year
It is stale the week after someone joins, leaves, or changes team
The check uses the live employee and organisation records
“Data quality” as a single percentage
Leadership cannot tell a missing phone number from a person who left and is still active
Serious, warning, and info - and you can open each one
HR opens profiles one by one
Duplicates and missing managers hide in a directory of hundreds
Duplicate emails, codes, and phones are grouped; missing managers are a list
Someone “cleans” records in bulk without looking
A wrong manager or exit date is now harder to undo
The page only points you to the screen. A person makes the change
Headcount that includes people who already left
Planning, billing, and leadership reviews start from the wrong number
Left-but-still-active and exit-before-joining are called out as serious
Each one is a real workplace mess, closed by looking at the records you already have.
Leave or attendance requests never reach anyone because the reporting line is broken.
Flags no manager, a person set as their own manager, an inactive manager, and loops in the reporting line.
Attendance marks late and absent against the wrong rules - or no rules at all.
Flags active people with no shift, and people with no attendance policy that applies to them.
Someone tries to apply leave and there is no balance for the current period.
Flags joined, active people who have no leave balance covering today.
Two people share a work email, employee code, or phone number.
Duplicate checks group the matching records so HR can see who clashes with whom.
Birthdays, emergency calls, and address letters fail because the profile was never finished.
Flags missing date of birth, phone, emergency contact, address, or employee code.
A department still has people, but the department is inactive or has no head.
Flags inactive departments with active people, missing department heads, and designations that do not belong to that department.
Someone left last month and still counts as active. Or their exit date is before they joined.
Those lifecycle conflicts are marked serious, with the people named.
Probation was supposed to end and nobody set a date, or the date has already passed.
Flags probation with no end date, and probation that is already overdue.
HR only hears about bad data when payroll, a manager, or an employee complains.
Open Data Health, refresh, and work the failing checks before the complaint arrives.
A report that only says “data quality is 82%” does not tell HR who to call this morning.
Each problem has a name, a count, and the people or departments behind it. HR can act the same day.
A broken reporting line or a person who left but is still active is treated as more urgent than a blank address. You can filter to what will actually stop work.
Changing a manager or an exit date is a human decision. The page shows the problem and opens the screen where you correct it.
After the record is corrected, refresh the check. The affected count is how you know the cleanup worked.
Shared emails, codes, and phone numbers are listed together so you are not hunting the second person by memory.
Bank details and tax numbers for payroll readiness are not part of this check. It looks at people, reporting, attendance setup, leave balances, and employment status.
A story you can explain on a pricing call. No empty tile on Starter that looks like a health check and does nothing.
Data Health is not on Starter - on purpose.
Same Data Health on both plans.
Teams usually discover messy employee records when leave, attendance, or headcount is already wrong. This page exists so that discovery can happen earlier.
A count without a name is a worry. A name plus the screen where you correct it is a morning’s work.
There is no button that changes a whole list of managers or exit dates at once. That keeps a health check from becoming a new source of mistakes.
Filter to serious items, or to one area such as reporting or attendance setup, and leave blank addresses for later.
Come back from the employee screen and refresh. The point is to watch the affected count drop, not to file a report and forget it.
The records stay where they already live. Data Health is the inspection. Core HR is where you correct the person.
Book a short walkthrough: open the score, filter to the serious gaps, open one person, and jump to the screen where HR actually fixes it. That usually settles it.