If you can't compute expected revenue from vehicle count, you cannot detect leakage. Most sites can't.
Parking revenue leakage is rarely theft. It is almost always the absence of a number to compare collections against — and in that absence, nobody can prove anything is wrong.
The one number that matters
If every vehicle is logged by plate with a timestamped entry and exit, expected revenue becomes computable: apply the tariff to the actual dwell distribution. Compare that to what was collected. The gap is leakage, and it can be reviewed daily instead of discovered during an annual audit.
Without a vehicle log, collections are taken on trust because there is no denominator. This is the actual condition of most manually-operated sites in India, and it is why leakage estimates in the sector vary so wildly — nobody is measuring the same thing.
The common leak paths
- Free-period abuse — validation stamps or grace windows applied to vehicles that did not earn them.
- Cash handling without shift-wise accountability, so a shortfall cannot be attributed.
- Manual overrides at the barrier, used routinely rather than exceptionally, and unlogged.
- Monthly passes that expired but still open the gate.
- Tariff misapplication — the wrong slab charged, usually in the customer's favour, at volume.
What fixing it looks like
Instrument first, then investigate. Plate-level logging gives you the denominator. Shift-wise cash accountability attributes any gap to a window and an operator. Logged manual overrides turn a routine workaround back into an exception someone has to justify. Automatic pass expiry closes the slowest and most persistent leak.
None of this requires accusing anyone of anything, which is precisely why it works. Most leakage is process drift, and process drift stops when it becomes visible.