Every Indian SME that runs payroll on a mix of Excel and paper registers is leaking money. Some of it is buddy punching, where one employee marks another as present, some of it is ghost employees, where salaries continue to flow months after someone left. Some of it is proxy attendance from a gate guard marking in a friend for a ₹200 bribe. Most HR heads know it is happening but cannot prove it, which is exactly how it survives.
The short answer is this. You stop ghost attendance and payroll fraud by removing the manual trust layer from attendance. Biometric hardware, face recognition, and geo-tagged mobile attendance together make it technically impossible for one employee to mark another or for a ghost name to stay on the payroll. Monthly audit rules and photo-proof attendance close the gap that every Excel-based system leaves open.
This guide breaks down the five most common types of payroll fraud in Indian SMEs, why manual systems cannot catch them, and what a modern attendance stack looks like. It is written for SME owners, HR heads, and operations managers across manufacturing, retail, construction, quick commerce, warehousing, and corporate offices.
Key Takeaways
TL;DR: Payroll fraud in Indian SMEs typically drains 3 to 8% of monthly payroll through buddy punching, ghost employees, proxy attendance, and inflated overtime. The MP Treasury flagged a ₹230 crore ghost employee case in May 2025. Biometric plus face attendance cuts this leak by up to 90% while closing the compliance gap that Excel leaves open.
What Is Ghost Attendance and Payroll Fraud in India?
Ghost attendance is when someone gets paid for time they did not actually work. Payroll fraud is the wider category that includes ghost employees (salaries paid to people who never worked at the company), buddy punching (one employee marking another), proxy attendance (a third party marking someone in), inflated overtime claims, and fake reimbursements.
In May 2025, the Madhya Pradesh Treasury flagged a ₹230 crore ghost employee fraud investigation after uncovering discrepancies in the state payroll system. The Commissioner of Treasury and Accounts sent notices to more than 6,000 Drawing and Disbursing Officers across the state. It is the largest publicly reported payroll fraud case in Indian government records in recent memory.
Private-sector figures are harder to track, but industry sources like OnGrid India place ghost-employee fraud among the top five payroll risk categories for Indian businesses, especially those running semi-manual payroll with heavy reliance on spreadsheets. The Indian Railways was reported to have lost around $187,600 to ghost employee fraud in earlier cases.
The global benchmark is stark. According to the Association of Certified Fraud Examiners (ACFE), payroll fraud schemes typically persist for 24 to 36 months before they are discovered, and businesses lose an estimated 5% of annual revenue to occupational fraud across categories. For an Indian SME with ₹10 crore in annual revenue, that is roughly ₹50 lakh a year disappearing into attendance leaks and shadow employees.
What Are the 5 Most Common Types of Payroll Fraud in Indian SMEs?
Most Indian SMEs leak money through the same five channels. Each one needs a different detection and prevention approach.
1. Buddy punching. One employee punches in or marks attendance for a colleague who is late, absent, or on leave. Most common in shift-based operations like a Surat garment factory, a Chennai retail showroom, or a Gurgaon construction site. At 200 employees on an average CTC of ₹8 lakh, even a 3% buddy-punching rate drains roughly ₹47,800 every month.
2. Ghost employees. People on the payroll who never show up or who left months ago but whose salary continues to credit into a bank account nobody ever closes. IndiaFilings reports that ghost employees most often appear in businesses without quarterly headcount audits or direct manager sign-off on payroll releases.
3. Proxy attendance. A third party, usually a security guard or a junior admin, marks attendance for someone in exchange for a small cash payment. Hardest to catch in warehouses, construction sites, and large manufacturing floors where the attendance register sits at a single entry gate.
4. Inflated overtime claims. Employees log OT hours they did not actually work, or managers approve excess OT for friends inside the team. A Hyderabad QSR chain audit in early 2026 found that 6% of its overtime hours across 14 outlets could not be reconciled against CCTV footage.
5. Unauthorised leave and late punches. Employees mark themselves present on days they skipped work, or adjust punch-in times to avoid late deductions. A Koramangala retail chain found that tightening late-punch rules alone recovered about ₹12,450 per month per outlet.
How Much Is Payroll Fraud Actually Costing Indian SMEs?
Hard numbers are difficult to pin down because most cases never surface. What we do know comes from two data points.
First, ACFE’s global finding that businesses lose about 5% of annual revenue to occupational fraud. For an Indian SME with ₹10 crore annual revenue, that is roughly ₹50 lakh a year. For a ₹50 crore manufacturer, it is ₹2.5 crore.
Second, Petpooja Payroll’s internal client observations. Across our 30,000+ client base, we see buddy-punching losses typically sit between 3 and 8% of monthly payroll in offices without face or biometric attendance in place. A Sarkhej manufacturing unit onboarded in late 2025 recovered roughly ₹3.2 lakh in the first three months after switching from a paper register to a biometric-plus-face attendance stack.
The scarier number is how long these frauds persist. ACFE places the average detection time for payroll fraud at 24 to 36 months under manual systems. That means by the time you notice, you have already paid out two to three years of leaked salaries.
What Are the Warning Signs of Ghost Attendance?
Before you audit your books, look for these operational signals. Any three together usually means money is walking out.
- Attendance punch-ins cluster within a 60-second window at shift start (one person punching for many)
- Monthly overtime claims are not supported by production output or CCTV footage
- Bank accounts receiving salary have not been linked to Aadhaar or updated in two or more years
- Employees rarely take casual leave yet never attend team off-sites
- Reimbursement claims come from the same three to four vendors nobody at the company can name
- Manager attendance approvals happen in a single Friday-evening batch without cross-verification
- Salary increments for “permanent” employees nobody has ever met in person
This is not an exhaustive list. It is the minimum signal set that should trigger a first-weekend attendance audit. Truein’s ghost employee detection guide walks through the audit checklist in more depth.
How Does Biometric and Face Attendance Stop Payroll Fraud?
The short answer: by making the punch physically attached to a unique body. You cannot fake a fingerprint, and you cannot beat a face-recognition system by sending a photo of your friend.
HRMantra’s biometric attendance guide and other industry sources place the buddy-punching reduction at up to 90% after switching from manual to biometric attendance. Face recognition with liveness detection closes the remaining gap.
Here is what each fraud type looks like before and after the switch.
| Fraud type | Manual system | Biometric + face |
|---|---|---|
| Buddy punching | Common, almost invisible | Technically impossible |
| Ghost employees | Persists 24 to 36 months | Exposed at first audit |
| Proxy attendance | Routine at gate-only registers | Blocked at device level |
| Inflated overtime | Easy to claim | Tied to actual punch-out time |
| Unauthorised late punches | Common | Flagged automatically |
The shift that biometric attendance creates in payroll accuracy is more than cost savings. It also closes the compliance gap that labour inspectors increasingly look for under India’s new wage code framework.
Mobile attendance with geo-tagging covers the field staff angle. A Petpooja mobile attendance app punch from outside a defined geofence gets flagged on its own, which stops field-force ghost punching from fake client sites.
How Does Petpooja Payroll Protect You From Attendance Fraud?
Petpooja Payroll was built around the exact leak points listed above, because its 30,000+ client base covers every fraud scenario at scale.
In-house biometric hardware. Comes with lifetime warranty on the device. A fingerprint punch cannot be faked by a friend, which closes buddy punching at the source. Devices sit at every entry point for factories, shops, warehouses, and offices.
Face recognition with liveness detection. The system checks for a live face, not a printed photo or a phone screen. This is what closes the remaining cases where biometric alone is beatable.
Geo-tagged mobile attendance. Field sales teams, service technicians, delivery riders, and site supervisors mark attendance through the mobile app inside a defined location boundary. Punches from outside the boundary get flagged for review. The Petpooja attendance management system treats all three attendance methods the same way on the backend.
Photo-proof attendance. For high-risk operations, the mobile app captures a live photo at punch-in. One visual audit removes doubt.
Payroll reconciliation dashboard. Monthly attendance rolls up to the salary calculation without manual input. Any employee with zero punches that month and an active salary line gets flagged before the payroll runs, not after.
Monthly audit reports on WhatsApp. The admin receives a daily and monthly attendance summary on WhatsApp. Anomalies show up in the report instead of waiting for a manual audit.
Across 30,000+ Petpooja Payroll clients we consistently see one pattern: the first biometric-plus-face attendance audit within the first 60 days of go-live recovers at least two to three times the monthly tool cost. The attendance discipline that this setup builds becomes the base layer for clean payroll everywhere else. It is the same logic a retail or F&B chain uses to prevent employee thefts, applied to attendance data instead of inventory.
Conclusion
Payroll fraud in Indian SMEs is not rare. It is routine, and most businesses are leaking 3 to 8% of monthly payroll without realising it. The MP Treasury ₹230 crore case and the Indian Railways losses are just the ones that made it to the headlines. For every one that did, a thousand smaller cases sit quietly inside Excel sheets and paper registers.
The fix is not complicated. It is biometric hardware at entry points, face recognition with liveness detection, geo-tagged mobile attendance for field teams, photo-proof punches for high-risk roles, and a monthly reconciliation dashboard that flags anomalies before the salary runs. One stack, one dashboard, one view of the truth.
The SMEs that build this stack stop losing ₹3 lakh a quarter to invisible leaks. The ones that do not keep paying, and the fraud persists the full 24 to 36 months that ACFE predicts before anyone notices.
Frequently Asked Questions
Payroll fraud is any act where an employee receives money they did not earn. Common forms in India include ghost employees, buddy punching, proxy attendance, inflated overtime claims, and unauthorised late-punch adjustments. A 2025 MP Treasury case uncovered a ₹230 crore ghost employee fraud, showing the scale this can reach inside large organisations.
Start with a physical attendance audit across a random 72-hour window using biometric or face recognition. Cross-check salary bank accounts against manager confirmations. Any employee with zero device punches in 30 days plus an active salary line is a strong ghost-employee candidate. ACFE data shows these schemes persist 24 to 36 months before detection under manual systems.
Yes, largely. Fingerprint biometric closes most of the gap, and face recognition with liveness detection closes almost all of it. Industry sources place the reduction at up to 90% once both layers are in place. Manual Excel systems have no equivalent control, which is why the fraud persists so long.
Petpooja Payroll runs biometric hardware, face recognition with liveness detection, geo-tagged mobile attendance, and photo-proof punches through a single dashboard. Any employee without punches in the month is flagged before salary runs, not after. Across 30,000+ clients, the first 60-day audit typically recovers more than the tool’s cost.
