AI improves on-time delivery by tracking every open order against its promised date, finding the order that will ship late while there is still time to save it, and getting the material, capacity or truck it needs from the person who owns it. On-time delivery (OTD) is the share of orders delivered on or before the date promised to the customer.
If late orders are costing you customers, Get your free Manufacturing AI Profitability Plan.
In this guide
What is on-time delivery, and how do you calculate it?
What is a good on-time delivery rate?
Why do orders ship late in 2026?
What does AI do about on-time delivery?
What does Morsa do for on-time delivery, step by step?
How much money does better on-time delivery make?
What has Morsa changed at a real plant?
How does a pilot on on-time delivery work?
What is on-time delivery, and how do you calculate it?
On-time delivery is the percentage of orders, or order lines, delivered on or before the committed date in a period.
OTD (%) = order lines delivered on or before the committed date / order lines due in the period x 100
OTIF (%) = order lines delivered on time and in full / order lines due in the period x 100
Four choices decide whether the number means anything:
Which date. Measure against the first date promised to the customer, not the latest rescheduled one. A plant that moves every date can score 100% while every customer waits.
Which unit. An order counts as late if any one of its lines is late, so order-level OTD is the stricter number. Line-level OTD shows where the misses are. Report the one your customer scores.
Shipped or received. Many customers measure receipt at their own dock. If you measure shipment from yours, say so.
The window. Some customers accept a few days early and none late. Use the window in the customer’s contract or scorecard.
A worked month: 200 order lines due in September. 168 left on or before the promised date, so OTD is 84%. Twelve of those 168 were short, so OTIF is 156 / 200 = 78%. Now say 22 of the 32 late lines had their dates moved before they were due: measured against the latest rescheduled dates, the same month reads 190 / 200 = 95%. A customer scorecard that counts OTIF against the first promised date says 78%.
OTD is a result, not a cause. Machine performance shows up in OEE, plan performance in schedule attainment, and repeat misses deserve a root cause analysis.
What is a good on-time delivery rate?
The honest answer is a range from the plants that publish theirs. Modern Machine Shop reported in September 2026 that Yeager Manufacturing, a Florida shop, saw on-time delivery slip to 70% during a period with many expedited orders, with 80% “a hard cap for the shop in better times.” Three months after adopting new scheduling software in February 2026, Zach Yeager reported 85 to 90%, trending toward 95%. MSP Manufacturing, an Indiana aerospace shop, holds 97%, according to Modern Machine Shop in June 2025, with owner and president Johnny Goode attributing most of the remaining 3% to delays in supplier deliveries.
PUBLISHED OTD, 2025 TO 2026
What is a good on-time delivery rate?
70%
on-time delivery at a Florida job shop during a run of expedited orders
Modern Machine Shop, September 2026
80%
the same shop's hard cap in better times
Modern Machine Shop, September 2026
85 to 90%
the same shop three months after adopting new scheduling software in February 2026, trending toward 95%
Modern Machine Shop, September 2026
97%
at an Indiana aerospace shop; most of the remaining 3% traced to supplier delays
Modern Machine Shop, June 2025
Single-shop figures reported in independent trade press. We found no free, public cross-industry OTD benchmark with a stated method.
Your customers’ scorecards set the real bar. Read their definition, date and window, then measure yourself the same way.
Why do orders ship late in 2026?
Because demand is outrunning supply and the slack is gone. ISM’s August 2026 report had supplier deliveries slowing for the ninth month in a row, production materials at an 84-day average lead time, and the Customers’ Inventories Index at 42.8%, which ISM reads as “too low.” Customers have little stock to absorb a late order. The Census Bureau’s July 2026 durable goods report showed unfilled orders of $1,599.9 billion, up in 24 of the last 25 months. ISM’s Backlog of Orders Index was 51.8%, still growing. In September, the New York Fed’s Empire State survey put its unfilled orders index at 5.9 and its delivery times index at 18.8, “significantly longer delivery times.”
A late order is usually late long before it ships. It is a purchase order that slipped, a job that fell behind on the schedule, a quality hold, or a truck that was never booked. Siemens’ True Cost of Downtime 2024 describes the chain: “if they miss their contractual obligations to buyers they are financially penalized. They have to add these costs to the penalties they impose on their suppliers for delivery failure. And so on, down the supply chain.” NIST’s survey of US manufacturers (published June 2020, 2016 data) found $100.2 billion a year in lost sales from maintenance-driven delays and defects, and 2.4 times more lost sales from maintenance-driven delays at the quarter of plants most reliant on reactive maintenance than at the quarter least reliant.
What does AI do about on-time delivery?
It turns every promised date into something watched, every day, from order entry to the dock. Each open order depends on a chain: material from a supplier, a slot on the schedule, a quality release, packing and a truck. AI reads each link where it is recorded, in the ERP, the schedule, the quality log, the dispatch plan and the messages between departments, and projects whether the order still makes its date.
When a link breaks, it works out the recovery while there is still time. The options are expediting the part, moving the job forward, splitting the shipment or booking a faster carrier. Then it gets it done by telling each owner in the channel they already use, and checks that it happened.
Johnny Goode of MSP Manufacturing described the other half of the job to Modern Machine Shop: “bad news up front is better than bad news the day the part’s supposed to be there.” When an order cannot be saved, AI tells sales early, with the reason and a new date, so the customer hears it before the ship date, not after.
What does Morsa do for on-time delivery, step by step?
Morsa runs the delivery-promise follow-up for you, working from your ERP, your customers’ dates and the channels your sales, planning and dispatch teams already use.
ONE ORDER, ONE PROMISED DATE
How an order at risk still ships on Friday
STEP 01
Signal
Track every open order line against its promised date, reading material, schedule, quality and dispatch status wherever each is recorded.
THE EXAMPLE
Monday: an order of 1,200 pieces promised for Friday; its coating supplier has not confirmed Wednesday
STEP 02
Consequence
Project the finish date from the weakest link, and the cost of missing: penalties, the customer's line, the account.
THEN
At the current date, coating lands Thursday evening; packing and the truck need it Thursday noon
STEP 03
Action
Within approved rules: expedite, resequence, split the shipment or rebook freight, and tell every owner. If the date cannot hold, tell sales now.
THEN
Purchasing gets the supplier to Wednesday; dispatch moves the pickup to Friday 14:00
STEP 04
Verification
Close on the proof of delivery or the shipping document, not on a reply.
THEN
Friday 14:10, the dispatch note is in the ERP; the order counts as on time
THE NEXT CHANGE STARTS THE LOOP AGAIN
Illustrative example. Steps are the ones Morsa runs on every open order.
Signal. Morsa reads open sales orders, purchase orders, the schedule, quality holds and dispatch plans from the ERP (SAP Business One, Business Central, Odoo or whatever your plant runs on) and the messages in WhatsApp, Teams or email.
Consequence. It projects each order’s ship date from its slowest link and ranks the orders at risk by what a miss costs.
Action. Inside rules the plant approves, it chases the supplier, proposes the resequence, asks quality for the release, or rebooks the truck. When the date cannot hold, it gives sales the reason and a new date to send.
Verification. It closes each order on the dispatch note or proof of delivery and reports OTD and OTIF by your customers’ definitions.
The upstream causes have their own pages: AI for supplier delays and AI for schedule attainment. For material specifically, see AI for material shortages.
How much money does better on-time delivery make?
For an illustrative mid-market plant with $100 million in annual revenue, roughly $400,000 to $800,000 a year from a 5 to 10 point gain in OTD. The plant is invented to show the arithmetic; each input is labeled as an assumption, so you can replace it with your own.
Input | Value | Source or assumption |
|---|---|---|
Annual revenue (Rev) | $100,000,000 | Illustrative mid-market plant |
OTD gain (G) | 5 to 10 points | Assumption: for example 85% to 95% |
Expedite and premium freight on a late order (e) | 2% of its value | Assumption |
Customer penalties on a late order (p) | 3% of its value | Assumption; 0% if your contracts have none |
Late-order revenue lost at the next renewal (k) | 10% | Assumption |
Contribution margin (m) | 30% | Assumption |
Formula: annual value = G x Rev x (e + p) + G x Rev x k x m
Revenue no longer shipped late: at 10 points, 10% x $100M = $10,000,000.
Expedites and penalties avoided: $10,000,000 x 5% = $500,000.
Business kept: $10,000,000 x 10% x 30% = $300,000.
At 10 points: about $800,000 a year. At 5 points: about $400,000.
With no penalty clauses, set p to 0: the result falls to $500,000 at 10 points and $250,000 at 5. The formula assumes late orders are spread evenly across revenue; if your late orders cluster on your largest customers, the value is higher.
What has Morsa changed at a real plant?
At J4S, a 120-person glass plant, on-time completion of operational commitments went from about 30% to about 75% in the first four weeks. That measures commitments, not an OTD rate. J4S runs Morsa and went live in two days with no new software, no migration and no training.
Anil Kohli, Production Head at J4S: “Our people don’t have to learn any new software. People just message the way they always have. Morsa coordinates all the messages in the background.” Read the J4S story.
How does a pilot on on-time delivery work?
It starts with the customers or product lines where lateness costs most. Connect. Morsa connects to the ERP and joins the channels your sales, planning and dispatch teams use. Configure. We agree the OTD definition your customers use, the baseline, the owners and the rules for what Morsa may do alone. Live. Morsa starts tracking every open order. At J4S the three steps took two days.
The pilot is free when no implementation work is needed. Otherwise there is a minimal implementation cost, refunded if the pilot shows no value. After the pilot, Morsa’s fee is a share of the value created, agreed once the numbers are in. It runs in Morsa’s cloud, your private cloud, or fully on-premise, including the AI models. Morsa works alongside the ERP and MES and replaces neither; how it sits beside your systems is in Morsa vs MES and ERP, what it runs across a whole plant is in our guide to autonomous manufacturing, and tools that build the plan are compared in production planning software.
To see what a higher on-time delivery rate is worth in your plant’s numbers, Get your free Manufacturing AI Profitability Plan.
Sources
Modern Machine Shop, Evan Doran, “Charting the Course Through Schedule Drift,” 8 September 2026, updated 14 September 2026 (Yeager Manufacturing). https://www.mmsonline.com/articles/charting-the-course-through-schedule-drift
Modern Machine Shop, Evan Doran, “Aerospace Shop Thrives With Five-Axis, AI and a New ERP,” 20 June 2025 (MSP Manufacturing; Johnny Goode quote). https://www.mmsonline.com/articles/aerospace-shop-thrives-with-five-axis-ai-and-a-new-erp
Institute for Supply Management, “Manufacturing PMI at 54.6%; August 2026 ISM Manufacturing PMI Report,” via PR Newswire, 1 September 2026. https://www.prnewswire.com/news-releases/manufacturing-pmi-at-54-6-august-2026-ism-manufacturing-pmi-report-302865127.html
US Census Bureau, “Advance Report on Durable Goods Manufacturers’ Shipments, Inventories and Orders, July 2026,” 26 August 2026. https://www.census.gov/manufacturing/m3/adv/pdf/durgd.pdf
Federal Reserve Bank of New York, Empire State Manufacturing Survey, September 2026. https://www.newyorkfed.org/survey/empire/empiresurvey_overview
Siemens, “The True Cost of Downtime 2024,” 181 interviews (vendor research). https://assets.new.siemens.com/siemens/assets/api/uuid:1b43afb5-2d07-47f7-9eb7-893fe7d0bc59/TCOD-2024_original.pdf
NIST, Douglas Thomas, “Economics of Manufacturing Machinery Maintenance,” AMS 100-34, June 2020 (2016 data). https://nvlpubs.nist.gov/nistpubs/ams/NIST.AMS.100-34.pdf
Morsa, J4S customer story (Morsa customer data). /customers/j4s
Changelog
24 September 2026: moved the worked calculation from an illustrative $50 million plant to an illustrative mid-market plant with $100 million in annual revenue and recomputed every result: about $400,000 to $800,000 a year, or $250,000 to $500,000 with no penalty clauses. The percentage inputs are unchanged. The step-by-step section now says what Morsa does and runs rather than what it builds, and the pilot section links to the autonomous manufacturing guide.
24 September 2026: first published. OTD and OTIF formulas with the four definition choices and a worked month, published OTD figures from 2025 and 2026, the ISM, Census and New York Fed demand and supply data, the four steps Morsa runs, a worked calculation for an illustrative $50 million plant, and the J4S result.
