Ask an operations director where their charging network loses money and the answer is rarely “fraud.” It is far more mundane: sessions that should have produced clean revenue but did not, and nobody noticed until month-end — if they noticed at all.
Leakage is a reconciliation problem
A charging session has to survive a surprising number of steps to become money in your account. The charger reports energy. A tariff is applied. A payment is authorized and captured. If the driver is roaming, a CDR is pushed to their provider. A partner statement is reconciled. A receipt is issued.
Every one of those steps can fail quietly:
- The meter value at the end of the session does not match the energy billed.
- A tariff version changed mid-month and the wrong one was applied.
- A payment service provider authorization never settled.
- A roaming CDR push failed and was never retried.
- A partner’s statement is missing sessions you know happened.
None of these throw an error a driver sees. They just produce a small gap. Multiplied across thousands of sessions, the small gaps are the leak.
Why the spreadsheet loses
The traditional answer is a month-end reconciliation: export everything, open a spreadsheet, and hunt for mismatches. This fails for three reasons.
First, it is too late. By month-end, a failed payment authorization may be unrecoverable and a roaming CDR may be outside the partner’s dispute window.
Second, it is unauditable. A spreadsheet has no chain of custody. When a number is corrected, there is no record of who corrected it, why, or what it was before.
Third, it does not scale. The spreadsheet that worked at 200 chargers is unworkable at 2,000.
What an exception queue does instead
The alternative is to treat failed reconciliation as a first-class operational object — an exception — surfaced continuously, not at month-end.
When a session fails a check, it lands in a queue with the reason attached: meter delta out of tolerance, tariff version conflict, unsettled PSP authorization, roaming reversal pending, energy delivered but zero billed. An operator reviews it, decides — reprice, write off, dispute, escalate — and the decision is recorded.
The difference is not just timing. It is that the work becomes:
- Continuous — caught while it is still recoverable.
- Assigned — a queue has an owner; a spreadsheet tab does not.
- Auditable — every resolution is backed by an evidence record, so the correction has a chain of custody.
How Sparqly handles it
Sparqly’s revenue assurance is a live exception queue. Checks run on a schedule, exceptions carry a write-off lifecycle, and a session can be repriced in place. Every action — repricing, write-off, dispute — produces an evidence record, so the chain from session to receipt to settlement to payout stays traceable.
Revenue assurance is not an accounting afterthought. For a charging business, it is the difference between the revenue you earned and the revenue you actually collected.