OTD is a promise kept (or broken), not a vague “fast delivery” idea
On-time delivery (OTD) is the percentage of deliveries that arrive on or before an agreed delivery date—or within an agreed delivery window. Procurement teams use it to judge reliability: does the supplier or carrier hit the date operations planned around?
The detail that matters is what “on time” means in your contract or PO. A delivery at 5:30pm might be “on time” for a site that books goods-in until 6:00pm, and “late” for a retailer with a strict 2:00pm slot. OTD only becomes useful when the rule is explicit.
How OTD is usually calculated (and where teams trip up)
Most organisations calculate OTD as: (number of on-time deliveries ÷ total deliveries) × 100. “On-time” can mean early-or-on-time, or strictly within a window. Both are defensible; they answer different questions.
A simple example
If you received 200 deliveries this month and 170 arrived within the agreed window, your OTD is 85%. If you count early deliveries as “on time,” that 85% might rise—yet your warehouse could still be dealing with congestion and unplanned receiving. That’s why the definition matters as much as the arithmetic.
Date vs. time window: “by Friday” is not the same as “Friday 10:00–12:00.”
Requested date vs. confirmed date: measuring against the buyer’s requested date can punish suppliers for dates they never accepted; measuring against confirmed dates can hide chronic over-promising.
Partial shipments: one PO delivered in three drops can inflate OTD if each drop is counted as a separate “delivery.”
Early deliveries: counting early as success can mask problems (storage, shelf-life risk, site congestion).
What counts as “arrived”: at gate, at dock, unloaded, booked into inventory—pick one and stick to it.
OTD vs. OTIF: why procurement cares about the difference
OTD only tells you about timing. It says nothing about whether the right items and quantities arrived. That’s why many teams pair OTD with OTIF (On Time In Full). A supplier can score 98% OTD while still short-shipping key lines and creating backorders. If your stakeholders complain “deliveries are on time but we still can’t build/serve customers,” you’re probably measuring the wrong thing—or only half of it.
Where OTD helps—and where it can mislead
OTD is most useful when late deliveries create real cost: production stoppages, premium freight, missed retail slots, idle labour, or service penalties. In those environments, a few percentage points can be the difference between stable operations and daily firefighting.
It can mislead when it becomes a target without context. Teams sometimes “fix” OTD by widening delivery windows, moving requested dates, or splitting shipments into multiple smaller deliveries. Your dashboard improves while your total cost and workload quietly rise.
Practical guidance for setting an OTD definition that survives real life
If you want OTD to drive better supplier behaviour (not better reporting behaviour), define it in operational terms and align it with how your sites actually work. The most common mistake is copying a generic KPI definition and then arguing about exceptions every week.
Write the “on-time” rule into the PO/contract: date or window, timezone, and what event counts as delivered (arrived, docked, unloaded, booked-in).
Agree which date is the baseline: requested date, confirmed date, or a mutually agreed “must arrive by” date for critical items.
Decide how to treat early deliveries: accept as on-time, on-time-but-early (separate bucket), or non-compliant for slot-controlled sites.
Define counting logic for partials and split shipments at PO line level if you need an honest view of performance.
Pair OTD with at least one companion metric (OTIF, damage rate, lead-time adherence, premium freight incidents) so you don’t reward the wrong behaviour.
A blunt take: OTD is a reliability metric, not a relationship metric
A supplier can be pleasant, responsive, and still miss dates. OTD gives procurement a clean way to separate “we like working with them” from “they meet commitments.” Just don’t pretend it’s objective if your own planning dates shift constantly. If buyers routinely change requested dates to match reality, OTD will look great while operations keep absorbing the chaos.