Measuring OTIF for Custom Orders

The OTIF metric measures the share of orders delivered on time and in full, counted against a promise window you define before the period starts. It is a stricter number than an on-time rate, because a late line or a short shipment fails the whole order. The practical rule for custom orders: fix the window, count the order as one unit, and report partials separately rather than letting them average into a score that hides them. This guide covers how to define the window, how to count failures, and where the metric misleads.

Most fulfilment dashboards show an on-time percentage and call it service level. That number stays high while customers wait, because it ignores the order that arrived late and incomplete.

Custom orders make the gap wider. A decorated garment has an artwork approval, a production slot and a ship date, and a delay at any of them lands on the same promise.

Why Does OTIF Expose More Than an On-Time Rate?

An on-time rate counts shipments. OTIF counts orders, and an order only passes when every line is complete and the parcel left inside the window.

That difference matters most in multi-item orders. A three piece order that ships two items on time scores as a pass on a shipment basis and a fail on an order basis, and the customer experience matches the second reading.

The metric also forces a definition of late. Without a written window, lateness becomes a judgement about whether a delay was reasonable, and that judgement usually lands on the side of whoever is reporting.

How Do You Define the Promise Window?

Start from the date the customer was given, not from the date production started. The window should end at the point you control, which for most dropshipping sellers is the handover to the carrier rather than the delivery scan.

Then decide whether the window is calendar days or business days. Calendar days match what a buyer reads on a product page, and business days match what a production calendar can plan against. Pick one and state it on every report.

Add the cut-off hour. An order placed at 23:00 behaves differently from one placed at 09:00, and a window that ignores the cut-off quietly moves work into the next day without recording it.

Write the definition down before the first measurement. A metric redefined after the result is a metric that always improves.

Report both bases when the order mix is mixed. An order basis is what the customer experiences, and a line basis is what production can act on.

How Should Failures Be Counted?

Count the order once, then tag the reason. A late artwork approval, a stock-out on one blank and a carrier pickup miss all fail the order, but they sit in three different functions and need three different fixes.

Split lateness into bands rather than a single flag. One day late and five days late are different service failures, and averaging them into one failure rate hides the second group.

Keep the partials visible. An order that shipped complete but late, and an order that shipped on time but short, should not land in the same bucket, because one is a scheduling problem and the other is a stock problem.

Our fulfillment team tags each OTIF failure with the function that owned it before the weekly review, which keeps the number from being discussed as a single blame-free percentage.

Where Does the OTIF Metric Mislead, and What Are the Limits?

It misleads when the window is set after the fact. A window adjusted to match performance turns the metric into a description of what happened rather than a promise to the customer.

It misleads on small volumes. A rate over twenty orders moves in five point steps, so a single busy week reads as a trend when it is noise.

It also hides severity. A 95 percent score can describe a hundred orders a day with five one-day delays, or a hundred orders a week with five week-long losses, and the two need different responses.

The last limit is scope. OTIF ends at handover, so it says nothing about transit, delivery or the condition of the parcel, which is where a large share of complaints begin.

Lead time sits upstream of the window and explains most of the variance in it. A blank that arrives late moves every downstream date, and a quality control hold adds a day without any order being scheduled badly.

Record the lead time you quoted next to the lead time you achieved. The gap between the two is what a buyer remembers, and it is the number that should feed the next window.

How Do You Use It Without Gaming It?

Publish the definition alongside the number, and keep the definition stable for a full quarter. Any change is recorded as a break in the series rather than a silent improvement.

Pair it with a transit measure so the handover boundary is visible. The window ends where your control ends, and the customer experience continues past it.

Review the failures by function, not by percentage. Which order problems sit outside fulfillment cover is a useful boundary check when a delay is being attributed to the wrong team.

Warehouse location changes the window more than most sellers expect. How POD warehousing cuts delivery time explains where a domestic warehouse shortens the controllable part of the promise.

Seasonal peaks distort the number, and the planning notes in Cyber Monday fulfillment cutoff planning show how a cut-off date changes what counts as on time.

When the metric is reported to a client or a marketplace, the supporting definitions usually need to match their wording. The scope and process of print order fulfillment covers the terms that recur in those conversations.

Suppliers can be compared on it, but only with the same window. How t-shirt fulfillment companies streamline apparel business sets out the operational differences that change what a promise should look like.

A stable OTIF metric is worth more than a high one, because it lets a seller promise a date and keep it. Measure the promise you make, count every order once, and read the failures by the team that owns them.

Frequently Asked Questions

Should OTIF be measured per order or per line?

Measure both. The order basis reflects what the buyer experiences, while the line basis tells production which item caused the failure.

Does OTIF include transit time?

It includes whatever the window says, and the OTIF definition should say it explicitly. Most sellers end the window at carrier handover, because that is the last point they control, and report transit separately.

What counts as in full for a custom order?

Every unit ordered, in the correct size and decoration, in one shipment unless a split was agreed. A substituted blank is not in full, even when the customer accepts it.

Is a 95 percent OTIF score good?

It depends on the window and the volume. Compare it against the same definition in a previous period, and check the severity of the failures before treating the number as a target.

How often should the metric be reviewed?

Weekly for the failures and monthly for the rate. Weekly review catches a drifting process, while a monthly rate is stable enough to compare across periods. Report the OTIF metric against the same written window in both, so a change in the score never comes from a change in the definition.

Related Articles

The Top Ecommerce Platform for Print on Demand Businesses

How Do You Compare T-Shirt Fulfillment Companies?

Discover the Best Companies for Custom Printing and Fulfillment