When Manhattan SCALE and your ERP disagree on inventory, the numbers are rarely wrong at random — the gap almost always traces to a specific break in the flow between the two systems. Here's how to isolate it methodically rather than reconciling line by line forever.
1. Rule out timing before assuming an error
SCALE and the ERP are often reconciled at different moments. If SCALE posts in real time and the ERP updates on a batch cycle, a snapshot taken mid-cycle will always look "wrong". Confirm you're comparing the two systems at a comparable point in time before treating a difference as a genuine discrepancy.
2. Check the interface between SCALE and the ERP
Inventory sync depends on messages flowing successfully in both directions. Look for:
- Failed or stuck messages: adjustments, receipts, or shipment confirmations that erred out and never posted to the ERP leave the two systems permanently out of step until reprocessed.
- Partial processing: a message that posted a header but failed on detail lines can create a subtle, hard-to-spot gap.
- Sequencing: out-of-order message processing (a shipment confirming before its receipt) can produce temporary or persistent mismatches.
3. Trace adjustments made in one system but not the other
Manual inventory adjustments, cycle-count corrections, or status changes made directly in SCALE — or directly in the ERP — that aren't configured to flow across are a classic source of drift. Identify where adjustments are being made and confirm each type is designed to propagate.
4. Compare units of measure and conversions
If SCALE tracks in one UOM and the ERP in another, an incorrect or missing conversion will make identical physical quantities appear as different numbers. Verify the conversion factors match on both sides for the affected items.
5. Check status mapping
SCALE and the ERP may classify inventory statuses differently (available, held, in-transit, quality). If held or in-transit stock is counted in one system's "on hand" and excluded in the other's, the totals will never match even when nothing is actually wrong. Confirm how each status maps across the interface.
6. Isolate with a single transaction
When the pattern is unclear, follow one transaction end to end: create it in SCALE, watch the message, confirm it lands in the ERP with the same quantity, UOM, and status. The step where it diverges is your answer.
The faster way
Cross-system inventory gaps are precisely what SupplyChain Assist helps with — describe the discrepancy and the systems involved, and it walks the likely break points and the checks for each, so you spend your time fixing the gap rather than hunting for it.