← Back to Field Notes

July 31, 2025 · The Bloomfield Team

Top 10 Mistakes Manufacturers Make Tracking OTD

Ask a manufacturer their on-time delivery rate and you will get a number between 85% and 98%. Ask how they calculate it and the confidence in that number evaporates. OTD is the metric customers care about most, the metric suppliers report most frequently, and the metric most manufacturers measure incorrectly.

A 2023 survey by APICS found that manufacturers who recalculated their OTD using the customer's original requested date instead of the internally revised date saw their rate drop by an average of 11 percentage points. The gap between the number you report and the number your customer experiences is where trust erodes.

1. Measuring Against the Revised Date Instead of the Original

The customer requests delivery by July 15. Your shop realizes it cannot make that date and negotiates a new date of July 22. The parts ship July 21. Your ERP shows on-time. The customer's system shows six days late. Both are technically accurate, but only one reflects the customer's experience. If your OTD calculation uses the most recent promise date, your number is artificially inflated. For a deeper look at delivery metrics, see our guide to production visibility.

2. Counting Partial Shipments as On-Time

An order for 500 parts is due July 15. You ship 400 on July 14 and the remaining 100 on July 22. Is this on-time? Most ERP systems record it as on-time because a shipment was made before the due date. The customer who cannot run their assembly line at full capacity until all 500 parts arrive would disagree.

3. Excluding Canceled or Rescheduled Orders

An order is running late. The customer reschedules or cancels before the original due date. If your OTD calculation excludes these orders, it removes the evidence of problems that forced the customer to change plans. Every canceled or rescheduled order should be investigated for root cause, even if it does not count against the OTD number.

4. Using Ship Date Instead of Delivery Date

Your parts leave the dock on July 14. The customer receives them July 18. Their PO said "delivery by July 15." You recorded on-time. They recorded late. If your customer measures delivery date and you measure ship date, your numbers will never align. The best practice is to track both and report the one your customer uses.

5. Not Weighting by Dollar Value

Your OTD shows 93% based on order count. But the 7% of orders that shipped late included your two largest customers and represented 22% of monthly revenue. An unweighted OTD treats a $500 order and a $50,000 order as equal. Revenue-weighted OTD tells you where late deliveries actually hurt the business.

6. Measuring Monthly Instead of Rolling

A monthly OTD number hides patterns. If you miss eight deliveries in the first week of the month and hit every one after that, the monthly number shows 82%. But the damage was concentrated in one week, probably caused by a specific bottleneck or scheduling failure. A rolling 30-day OTD with weekly granularity reveals patterns that monthly reporting conceals.

7. Not Tracking Root Cause

Recording that a shipment was late is step one. Recording why it was late is where improvement starts. Was it a material delay, a quality hold, a scheduling conflict, a capacity shortage, or an outside processing vendor who missed their lead time? Without root cause data, the OTD number is a score without a game plan.

Shops that categorize every late shipment into five or six root cause buckets can identify systemic issues within 90 days. Material delays causing 40% of late shipments tells you to fix your material planning. Scheduling conflicts causing 35% tells you to fix your capacity model. Without the categorization, every late shipment looks like a one-off.

8. Ignoring Internal OTD Between Departments

External OTD measures when the customer gets their parts. Internal OTD measures whether each department hits its handoff target. If machining finishes on time but grinding adds two days, the late shipment looks like a grinding problem. If machining finishes one day late and grinding absorbs the delay through overtime, the numbers show on-time but the overtime cost is hidden.

Tracking internal OTD at each production stage identifies which departments are creating the delays and which ones are hiding them through heroic effort.

9. Setting the Wrong Targets

A shop sets an OTD target of 95% and celebrates when they hit it. Meanwhile, their largest customer requires 98% OTD to maintain preferred vendor status. The target should come from the customer's expectations, not from an internal goal-setting exercise. Different customers may have different thresholds, and your most valuable customers define the standard.

10. Not Connecting OTD to Leading Indicators

OTD is a lagging indicator. By the time you know a shipment is late, the damage is done. The leading indicators that predict OTD are: schedule adherence (are jobs starting when planned), material availability (is material on hand when needed), first-pass yield (are parts passing inspection the first time), and outside processing lead time (are vendors delivering on schedule).

Shops that track these four leading indicators weekly can predict OTD problems five to seven days before the due date. That is enough time to reallocate resources, expedite a process, or communicate proactively with the customer. Fixing the leading indicators fixes OTD. Tracking OTD alone tells you about problems that have already happened.

Getting the Number Right

An honest OTD number, measured against the original customer-requested date, including partial shipments as late, weighted by revenue, with root cause tracking on every miss, is almost always lower than the number currently reported. That is the point. The honest number is the one that drives improvement. The inflated number is the one that lets problems persist while the dashboard looks acceptable.

The manufacturers who fix their OTD tracking first and their OTD performance second are the ones whose customers trust their delivery dates. Trust in delivery is what earns repeat business, preferred vendor status, and the long-term contracts that stabilize revenue. The metric has to be real before it can be useful.