Interloop Editorial Team
Most HR teams don't think of themselves as running a data business. But every day, they create and change data that the whole company depends on: who works here, what they're paid, who they report to, where they sit, what they're entitled to and what they can access.
When that data is right, nobody notices. Payroll runs, approvals reach the right manager, and reports add up. When it's wrong, the effects show up everywhere, often far from HR. A salary goes to an old bank account. A leave request lands with a manager who left last year. A board report shows a headcount that finance doesn't recognise. An ex-employee still has access to a shared drive.
These problems rarely look dramatic one at a time. That's what makes them expensive. This piece looks at how employee data gets messy, what it really costs, and how to clean it up and keep it clean, using Zoho People as one example of how a single system helps.
Key insights
- Bad data is the norm, not the exception. Researchers Nagle, Redman and Sammon analysed 75 data quality assessments across organisations. On average, 47% of newly created records had at least one critical error.
- Very few companies meet the bar. In the same research, only 3% of the data quality scores were rated acceptable.
- Exits are a weak spot. Research cited by Zoho from Wing Security found former employees still had access to sensitive company data in 63% of organisations.
- India's privacy law raises the stakes. The Digital Personal Data Protection Rules were notified in November 2025, with most obligations taking effect after an 18-month phase-in, around May 2027. Every employee, candidate and dependent whose data you hold is covered.
- Penalties are significant. Under the DPDP Act, failing to take reasonable security safeguards can attract a penalty of up to ₹250 crore.
How employee data gets messy
Nobody sets out to keep bad data. It drifts there, one small shortcut at a time. These are the patterns we see most often:
Too many copies. The employee master sits in a spreadsheet. Payroll has its own list. Attendance has another. IT has a directory for email and system access. Each one is updated at a different time, by a different person, so they slowly drift apart.
Changes that don't travel. Someone is promoted, transfers to another city or changes manager. HR updates one system. Nobody tells payroll, IT or finance until something breaks.
Free-text everything. One record says "Bangalore", another "Bengaluru", a third "BLR". Designations are typed by hand, so "Sr. Executive" and "Senior Executive" count as two different roles in every report.
Missing fields. Emergency contacts, PAN numbers, bank details or reporting managers are left blank during a rush to onboard, and nobody goes back.
Exits that aren't fully closed. The person leaves, final settlement is processed, but their system access, licences and records are left as they were.
Growth. What worked for 40 people in one office breaks at 400 people in four cities. Most companies don't notice the moment it breaks. They only see the symptoms later.
Where it hurts
The costs of messy employee data show up across the business. Most of them are hidden inside other problems, which is why they're easy to ignore.
| Area | What goes wrong | The real cost |
|---|---|---|
| Payroll | Wrong bank details, outdated salary structures, missed attendance or leave data | Rework, late salaries, unhappy employees, and occasionally money that's hard to recover |
| Compliance | Missing statutory details, records that can't be produced in an audit | Penalties, notices and time spent answering regulators |
| Data privacy | Ex-employees with live access, personal data stored in uncontrolled spreadsheets | Security incidents and, under India's DPDP Act, significant penalties |
| Approvals and workflows | Leave and expense requests going to the wrong manager | Delays, frustration and approvals given by the wrong person |
| Reporting | Headcount that doesn't match finance, people counted twice | Leaders stop trusting HR numbers and make decisions on gut feel |
| Software costs | Licences still assigned to people who have left | Paying every month for accounts nobody uses |
The reporting cost is the one we'd underline. Once leaders find one wrong number in an HR report, they start doubting all of them. Rebuilding that trust takes much longer than fixing the data did.
Privacy changes the conversation
For companies in India, there's a new reason to act. Under the DPDP Act, the company is responsible, as a data fiduciary, for the personal data it holds about employees, candidates and their families. That means knowing what data you hold, why you hold it, who can access it, how long you keep it, and how you protect it.
That's very hard to answer when employee data is spread across a dozen spreadsheets and inboxes. With the main obligations expected to apply from around May 2027, cleaning up employee data is no longer just good housekeeping. It's preparation for compliance. Companies operating in the UAE have their own personal data protection rules to consider as well.
One source of truth
The single most effective fix is structural: decide where the master employee record lives, and make every other system take its data from there. Payroll, attendance, IT access and reporting should read from the same record, not keep their own versions.
A dedicated HR system makes this much easier. Zoho People, for example, holds the core employee record with fields you can configure, and lets you make key fields mandatory and restrict others to drop-down values, so "Bengaluru" is always spelt the same way. Changes such as promotions, transfers and manager updates can trigger workflows, so the people who need to know are told automatically. Role-based permissions control who can see sensitive fields such as salary or bank details. When Zoho People is connected to the rest of the Zoho HR suite, such as Zoho Payroll, the same record drives salary processing, rather than a separate list that has to be kept in sync by hand. Employees can also update their own contact details through self-service, which keeps data fresher than any annual form.
The system is an enabler, not the cure. A clean system still needs someone who owns the data and a few simple rules about how it changes.
How we suggest cleaning it up
- List every place employee data lives. HR files, payroll, attendance, IT directories, finance sheets, managers' own trackers. You'll probably find more than you expected.
- Choose the master. Decide which system holds the official record for each field, and stop maintaining copies elsewhere.
- Define the core fields. Agree the 20 to 30 fields that matter most, how each is formatted, and which are mandatory.
- Clean in one focused push. Compare sources, fix mismatches, fill gaps and remove duplicates. Ask employees to verify their own details where you can.
- Name a data owner. One person in HR should be accountable for the quality of employee data, with clear responsibilities for others who update it.
- Make changes travel automatically. Set up workflows so that joiners, movers and leavers update every connected system, including IT access.
- Close exits properly. Build a leaver checklist that includes revoking system access and licences on the last working day.
- Audit every quarter. Run a short check of mandatory fields, duplicate records and active accounts for people who have left.
- Map your privacy obligations. Record what personal data you hold, why, who can see it and how long you keep it. Use this to prepare for DPDP compliance.
How you'll know it's working
- Share of employee records with all mandatory fields complete
- Number of duplicate or conflicting records found in each quarterly audit
- Payroll corrections caused by employee data errors
- Days between an employee's exit and the removal of all system access
- Number of times HR and finance headcount figures differ
- Time taken to produce a standard headcount or attrition report
When leaders stop asking "are you sure this number is right?", you'll know the data is in good shape.
Interloop. Orchestrating business — today, and tomorrow.
