When a trend becomes a decision: supplier OTIF drift

A key supplier's on-time-in-full slides from 98% to 88% over six weeks. The agent watches three periods, then recommends acting — before the line stops.

The situation. A single supplier feeds a critical machined component to two assembly lines. Their contractual OTIF floor is 95%, and for two years they have lived at 97–98%.

The signal. The Supplier Performance agent watches OTIF per supplier against the SLA with a persistence rule — one bad week is noise; three deteriorating periods is a trend. This supplier slides 98% → 93% → 88% over six weeks, and lead-time variance widens alongside. Systems consulted: ERP (PO receipts, promised vs actual dates), supplier scorecard, and inbound schedules from the TMS.

The decision. Doing nothing carries a priced risk: at the current trajectory, coverage on the two fastest-moving SKUs breaks in about three weeks, and a line stoppage idles roughly $22,000 per day — call it $44,000 of realistic exposure. The agent recommends a split response: expedite one open PO (premium freight ~$9,000) to rebuild the buffer, shift 30% of the next two months’ volume on those SKUs to the qualified alternate source, and open a corrective-action request with the primary rather than quietly punishing them.

The write-back. Expedite flag and revised routing on the open PO in the ERP; new split-award POs to the alternate; the corrective-action case logged against the supplier record with the evidence attached.

The outcome.

Metric	Result
Line stoppages	0 (exposure ~$44,000 avoided)
Premium freight spent	$9,000
Component availability OTIF	Back ≥95% within 4 weeks

The takeaway. Write down, in advance, how much drift you will tolerate before it becomes a decision. The trend will not tell you.

Representative scenario; suppliers and figures are synthetic. SCOR is ASCM’s framework.