Here is the mechanism, and it’s almost innocent. An expense comes in ₹800 over the hotel cap. The approver is busy, it’s a good employee, it’s a rounding error against the budget — approve. Repeat that a few thousand times a year across every category, and the “rounding errors” add up to real money that never shows on any report as a problem, because every single one was approved.
The leak isn’t caused by bad people. It’s caused by two structural gaps: approvers can’t see the policy breach at the moment they click, and nothing sizes the total until year-end — by which point it’s just “the budget.”
Where the budget goes — the leak, made visible. Seven percent doesn’t sound like a crisis until you multiply it by an annual T&E budget. On ₹5 crore of travel, that’s ₹35 lakh a year. Illustrative proportions.
Put the leak on a slide the board understands
- ~7% — Of T&E spend that clears while out-of-policy (the leak rate)
- ₹35 L — Annual leak on a ₹5-crore travel budget (illustrative)
- seconds — An approver spends on the average claim (why breaches slip)
Out-of-policy spend — before vs after enforcement. The biggest drop comes not from rejecting more, but from flagging the breach before it reaches an approver. Illustrative outcome.
Move the check to where the decision is
Fix 1 — Check at submission, not after
Flag the over-limit hotel or missing GST the moment it’s entered — to the employee, before it becomes an approver’s awkward call. Most breaches self-correct.
In MyVyay Real-time policy checks at capture
Fix 2 — Auto-approve only what’s clean
Let in-policy claims clear themselves and route only the exceptions to a human — who now sees the breach highlighted, with context, and actually decides.
In MyVyay Role-based auto-approval + exception routing
You can’t approve your way out of a leak. You have to make the breach visible at the exact second someone clicks — or it becomes next year’s baseline.
The 7% isn’t a morale problem or a training problem. It’s a timing problem: the policy check happens too late to matter. Move it to the moment of submission, auto-clear the clean majority, and the leak shrinks to a rounding error you can actually defend.
See it on your own numbers: book a 20-minute demo or start free and run this playbook against your real spend data.