Where revenue quietly leaks
Leakage is rarely a single dramatic failure. It is a hundred small mismatches between what was provisioned, what was billed and what was actually collected.
In twenty-three years across telecom, DTH and OTT, I have almost never found revenue leakage in a place anyone expected. It is not usually fraud, and it is rarely one broken system. It is the quiet distance between three versions of the truth: what the network provisioned, what the billing system charged, and what the bank finally received.
The first place I look is the boundary. Every handover between systems — provisioning to billing, billing to payments, distributor wallet to customer account — is a place where records can be dropped, duplicated, timed differently, or rounded inconsistently. A mismatch of a rupee on a single transaction is noise. The same mismatch across thirty-five lakh tertiary transactions is a reporting problem and a real financial one.
The second place is the exception queue nobody owns. Most organisations already generate exceptions; far fewer have an accountable path from exception to root cause to closure. An exception report that is produced daily and read weekly is a record of leakage, not a control against it.
The third place is timing. Cut-offs, posting dates and settlement lags create differences that look like leakage and are not, and hide differences that are leakage and look like timing. Separating the two is most of the work in any reconciliation, and it is why period definitions must be agreed before the numbers are.
What actually reduces leakage is unglamorous: define the expected flow end to end, reconcile at every boundary rather than only at the end, classify every exception by cause rather than by amount, and measure how long exceptions stay open. Once causes are classified, the pattern becomes obvious — a tariff configured wrong, a promotion not expiring, a payment channel posting twice on retry.
The reason this matters beyond the finance function is that the same discipline produces trustworthy reporting. When reconciliation explains itself, forecasting improves, the close shortens, and management stops debating whether a number is right and starts deciding what to do about it.