Why your BC agents fake GPS location — the four patterns and how to stop each one

Every branch head running a Business Correspondent (BC) network has had the same conversation with a supervisor: "The check-ins look fine on the dashboard, but the customer says nobody came." Sometimes it is a one-off. Often it is a pattern. And the reason people fake location is not always what you assume.

Here are the four patterns we see in real MFI, NBFC, and BC operations across India — what causes each, what it looks like on the dashboard, and the specific countermeasure that stops it.

Pattern 1: The lazy officer

What happens: The officer is at home or the branch, but marks a check-in "at the borrower location." They plan to actually visit later — or not at all if the borrower calls to say they will pay via UPI.

Why it happens: Path of least resistance. The daily quota looks met. If the borrower pays anyway, nobody notices.

On the dashboard: The check-in GPS pin is nowhere near the borrower's address. Often clustered around the officer's home or the branch office. Selfie (if captured) shows an indoor background instead of a shop or house.

How to stop it:

• Proximity check: reject any check-in where distance to the known borrower location is > 200m
• Selfie background: mandatory selfie. Supervisor reviews random samples.
• Address quality: audit borrower addresses in your CRM. A vague address like "near temple" makes proximity checks impossible.

Pattern 2: The mock GPS ring

What happens: A group of officers coordinate. One of them figured out that a rooted Android phone can spoof GPS with a free app. They share the technique. Suddenly the whole team's check-ins look perfect.

Why it happens: Peer pressure to match a top performer's numbers, plus the incentive structure rewards visit count over visit quality.

On the dashboard: Check-ins look too perfect — always exactly at the borrower's GPS coordinate, always within seconds of scheduled time. GPS accuracy is unnaturally low (5m or below every time). No natural jitter.

How to stop it:

• Mock-GPS detection: Android tells the app when the location is mocked. Reject or flag any such check-in.
• Accuracy signature: real GPS has a distribution of accuracy values. Flag officers whose average accuracy is suspiciously low.
• Timestamp variance: real visits do not happen at exactly the same second every day. Flag officers with too-clean timing.

Pattern 3: The batch fraud

What happens: Officer visits nobody all morning. At lunch, they open the app and rapid-fire eight check-ins across the borrower list — some real, some spoofed.

Why it happens: Officer covered actual work in the afternoon or intends to catch up "on the weekend." Reporting quotas need to be met daily.

On the dashboard: Multiple check-ins within a short window. Physically impossible travel time between two consecutive check-ins (e.g., check-in at village A and 3 minutes later at village B 10 km away).

How to stop it:

• Travel-time sanity check: calculate implied speed between consecutive check-ins. Speeds > 80 km/h in rural areas are almost always fake.
• Minimum gap enforcement: reject a second check-in within 5 minutes of the first, unless the officer is at the same location.
• Daily distribution alerts: dashboard flags officers where all check-ins happened in a two-hour window or less.

Pattern 4: The ghost supervisor

What happens: Not the officer — the supervisor fakes visits from their laptop. They approve fabricated reports, sometimes because their own performance bonus depends on branch productivity, sometimes to cover an officer's actual absence.

Why it happens: Incentive alignment failure at the supervisor level. Also: the supervisor's boss trusts the aggregate number, not individual check-ins.

On the dashboard: The check-in came from a desktop-shaped device fingerprint (large screen, no motion sensors). Or the officer's phone was recorded elsewhere at the same time (device location logs).

How to stop it:

• Device binding: check-ins must come from the officer's registered device
• Single-device policy: one active session per officer at a time
• Field-only mode: block check-in submissions from IPs known to be corporate/branch networks

The meta-fix: incentive structure

All four patterns share one root cause: the incentive rewards check-in count, not check-in quality. If your monthly review says "Rajesh had 25 visits per day," Rajesh will produce 25 visits per day. Some will be real; the rest will be whatever it takes.

The fix is to change what you measure:

• Reward verified visits, not reported ones
• Reward collection amount, not visit count
• Reward verified callback confirmation rate (random borrower calls that confirm the visit happened)
• Publicly celebrate officers who reject impossible route plans instead of faking through them

You will not eliminate fraud with software alone. But software gives you the evidence to change the incentive structure — and that is what actually fixes the problem long-term.

How CollectRoute helps

CollectRoute implements every countermeasure above by default:

• Mandatory GPS + selfie + timestamp on every check-in
• Mock-GPS detection built into the mobile app
• Proximity threshold (200m) with per-tenant override
• Rapid-check-in and travel-time sanity checks on the dashboard
• Device binding + single-active-session enforcement

See the microfinance vertical page for how it fits an MFI or NBFC workflow, or start a free trial — takes about 10 minutes to set up your first branch.