A wholesale invoice can be completely correct and still tell a commercial team very little about what changed. It states what the supplier expects to be paid. It rarely explains which traffic created the amount, which rates were applied or why this month looks different from the last.
That is where reconciliation starts. Matching one total to another may confirm that two systems agree, but it does not explain the result. For an MVNO, the commercial work sits below the invoice line, where millions of usage events become rated charges and eventually a month-end number.
The total is the end of the calculation
Wholesale invoices are built for settlement. They summarise traffic by period, service, destination or product. That compression is necessary, but it hides the commercial story. One total can contain higher volume, a different destination mix, roaming, late files, corrections, minimum charges and several rate treatments.
If a bill is 12 per cent higher than last month and the only explanation is that usage went up, you have spotted a variance. You have not reconciled it. The increase might be expected growth, a small group of expensive subscribers, a rate change or a classification difference. None of those automatically means the invoice is wrong.
Commercial confidence comes from being able to explain the number, not merely reproduce the total.
Keep the route back to usage
A useful reconciliation view keeps the chain intact. The amount charged sits at one end. Behind it are the supplier’s rating output, the commercial terms and the activity to which those terms were applied. The operator’s own usage data provides another view of the same traffic.
The job is to compare like with like and keep enough context to answer practical questions. Which service, period, destination, subscriber or SIM contributed to the charge? What volume was rated? Which rule was used? Where do the two sides differ in completeness, classification or value?
A red or green status cannot do that on its own. Finance may need to explain an accrual difference. Commercial may need to test a tariff against its cost base. Operations may need to find whether a mismatch began in mediation, rating or subscriber activity. Each needs a way into the detail, not another summary.
A difference is not automatically an error
Two systems can produce different answers for legitimate reasons. Cut-off times vary. Reference data drifts. Events are grouped differently. Rounding creates small movements, while late files create larger ones. The invoice is innocent until the data proves otherwise, which is less dramatic than it sounds and considerably more useful.
Once the difference is described properly, the next step becomes clearer. Some items need no action. Some need an internal correction. Others justify a precise supplier question, supported by the affected period, service, rate and underlying activity. That is a much stronger starting point for dispute preparation than a spreadsheet with one large number circled in red.
Materiality matters too. An operator does not need a committee meeting for every rounding difference. It does need agreed tolerances, a view of repeated small discrepancies and a route to the underlying activity when a threshold is crossed. Otherwise minor leakage can become a permanent feature of the bill simply because no single line looks exciting.
Telara’s reconciliation work is built around this route from charge to rated usage and supporting activity. Whatever system an operator uses, the principle holds: reconcile the explanation, not just the total.