Revenue leakage is a data problem, not an accounting problem

Revenue leakage has an image problem. The phrase conjures fraud, embezzlement, a bad actor with a hand in the till — something for the auditors and maybe the lawyers. That image is mostly wrong, and the wrongness is expensive, because it sends companies looking for villains when they should be looking for seams.

A systematic review that synthesized 1,784 academic articles on the subject lands on a quieter definition: revenue leakage is the money an organization has legitimately earned but never collects, lost not to crime but to fragmented processes, conceptual ambiguity, and gaps no one is watching. The losses are real and recurring. The cause is almost always structural.

The four places money actually leaks

The most useful recent work surveyed 210 practitioners across transaction-intensive industries — telecommunications (22.4%), e-commerce and retail (21.0%), healthcare and insurance (18.6%), and financial services (17.1%) — to ask not whether leakage happens but where. The answer came back as four detection constructs, and it’s worth noting what they have in common:

Pricing compliance — is the price that was charged the price that was contractually agreed? Discounts that outlive their terms, rate cards that drift from the contract, promotions that never switch off.

Authorization integrity — was every transaction that should have been billed actually authorized and captured? Services delivered but never invoiced, entitlements granted without a charge.

Temporal anomaly — does the timing line up? Charges that fire late, recur when they shouldn’t, or fall into the wrong period.

Adjustment behavior — are credits, refunds, and manual corrections within normal bounds, or is someone (or some script) quietly adjusting revenue away?

Look at that list again. Not one of them is theft. Every one is a question about whether two records agree: the contract versus the invoice, the service delivered versus the service billed, the charge versus the calendar, the adjustment versus the norm. Leakage is what happens in the space between records that were supposed to match and didn’t.

Why finance can’t fix it alone

This is why throwing more accountants at the problem rarely works. The general ledger is a destination. By the time a number reaches it, the transaction that produced it is over, the context is gone, and the discrepancy — if anyone notices — is a forensic exercise against stale data. Finance is structured to record what happened, not to catch what should have happened and didn’t.

The leak lives upstream, at the integration points where the CRM hands off to billing, where billing hands off to the ledger, where the delivery system confirms a service that the invoicing system never hears about. Those are data-engineering boundaries, not accounting ones. The reason the money escapes is that the systems on either side of the seam were never reconciled at the moment the transaction crossed it — exactly the same blind-handoff pathology that loses margin in physical distribution, only the cargo here is a billing record.

The fix is reconciliation at the seam

The emerging answer treats leakage detection as a continuous data problem rather than a periodic audit. Banking offers the clearest template: the lesson from revenue-assurance work in that sector is that you fight leakage by integrating the data — pulling the contract, the transaction, the authorization, and the adjustment into one reconciled view, then running detection continuously against it rather than sampling after the fact.

In practice that means a few principles:

Reconcile transaction-level, not summary-level. Leakage hides in individual records — the one contract whose discount never expired, the batch of services that slipped past billing. Aggregate reporting averages it away; you have to check at the grain where the discrepancy lives.

Make the check continuous. A quarterly audit catches a quarter’s worth of leak after it’s already gone. A standing reconciliation that flags the mismatch the day it happens turns a forensic recovery into a same-day correction.

Watch all four seams, not just pricing. Most leakage programs start and stop at pricing compliance because it’s the most visible. The survey’s other three constructs — authorization, timing, adjustments — leak just as reliably and get watched far less.

The reframe that pays

The companies that recover leaked revenue aren’t the ones with the strictest auditors. They’re the ones that stopped treating leakage as a finance failure to be caught after the fact and started treating it as a data-integrity property to be enforced in real time — the same posture good engineers bring to any system where two sources of truth have to agree.

Money rarely walks out the front door. It seeps through the seams. And seams are an engineering problem.