Compare sources only after identifying what each measures, its population, time basis and adjustments. Build a signed bridge for legitimate differences and isolate unresolved exceptions. Do not force agreement by changing values when booked activity, recognised revenue and collected cash answer different questions.
Establish whether the sources make the same claim
Two reports can use the same word sales while measuring different events. One may record signed orders, another issued invoices and another money received. Define the unit, population, currency basis and cut-off for each before comparing values. Agreement is meaningful only when the concepts should align. A smaller cash total than booked value can be a normal result of payment terms rather than a collection-system defect.
Create a comparison specification identifying the finest shared grain available. If both sources carry invoice identifiers, reconcile at invoice level before summarising. If one source is aggregated, state which differences cannot be traced further. Include cancelled orders, partial payments, credits and excluded business units in the scope review. An unsupported adjustment to make totals equal can conceal a genuine gap and destroy the evidence needed to resolve it.
Build a bridge with non-overlapping components
Start with one source total and add or subtract explicitly defined reconciling items to reach the other. Each component should have a sign, value, evidence and owner. Classify a record once in the bridge unless a documented decomposition requires separate amounts. Timing differences and scope differences should remain distinguishable because they have different expected resolution paths. Keep an unexplained residual rather than inventing a balancing adjustment.
Check whether converted currencies, tax inclusion or rounding contribute to the gap. Use a common basis where that is justified and retain the originals. Compare source arrival dates so a stale export does not masquerade as a conceptual difference. A bridge can explain why totals differ without proving either source is fully accurate. Its conclusion should state what reconciled, what remains unresolved and which uses are supported.
Investigate the residual rather than erasing it
Prioritise unresolved differences according to decision impact and recurrence. A one-off timing item may clear next period; a persistent unmapped source may require a structural fix. Set expected clearing dates and revisit aged items. Do not repeatedly carry forward a temporary label without evidence that the original explanation still applies. Preserve prior bridges so a reviewer can see whether the same exception has merely changed name.
The framework linked below supplies general consistency background. The worked bridge is an original fictional example and is not an accounting reconciliation standard. Share definitions with the report's readers, particularly when they regularly compare dashboards from different teams. That explanation can prevent unnecessary disputes while leaving genuine inconsistencies visible and assigned to someone who can investigate them.
For recurring bridges, keep a stable component dictionary. Renaming an unresolved item each month can obscure its age and make a persistent source problem appear to be a succession of unrelated small differences.
Fictional orders and collections: a bridge with a residual
A fictional scenario to explain the method, not a customer result or a market benchmark.
A fictional order system reports 20,000 units of booked value for June. The collections extract reports 14,000 received during June. Review finds 8,000 of June orders still unpaid, while 2,000 of June receipts relate to May orders. Starting from bookings, 20,000 minus 8,000 plus 2,000 equals 14,000, explaining the difference under this simplified scope.
A second source version shows collections of 13,800. The existing bridge now leaves a 200-unit residual. The analyst does not hide it in timing merely because timing already explains most of the gap. They identify affected payment references and retain the unresolved amount in the publication note until evidence supports a classification.
These invented totals deliberately separate event concepts and an unexplained exception. They do not establish revenue recognition, tax treatment or the correctness of either underlying system. The bridge makes a comparison reviewable while preserving the limits of available evidence.
TAKE THIS INTO YOUR NEXT CONVERSATION
A practical next-step checklist.
- Define each source’s concept, scope and clock.
- Use shared identifiers where available.
- Build signed, non-overlapping bridge components.
- Retain residuals with owners and review dates.
Common questions.
Should cash collected equal orders booked in the same month?
Not generally. Payment terms, prior-period receipts, cancellations and other scope differences can create legitimate gaps.
Can a bridge prove both sources are correct?
No. It explains identified differences under stated definitions; common errors or unsupported source values may remain.
Further reading & source context.
These primary references provide background. They are not evidence for the fictional example above and do not endorse Accuralytix.
Found an error or a missing qualification? Send an editorial correction.
Check my data readiness.
Use a practical tool to turn this idea into a clearer starting point.
Check my data readiness