Blog · Board and management reporting · Supply chain
OTIF, fill rate and on-time delivery are three delivery measures that are reported as if they were one. On-time delivery tests the date only. Fill rate tests the quantity only, and comes in line, case and order versions. OTIF requires both on the same order. This page sets out the definitions, computes all of them on the same ten order lines, shows why OTIF is always the lowest, and explains which one a retailer's scorecard and chargebacks are based on.
Three measures of the same deliveries. They test different things, and the differences explain most arguments between a supplier's report and a customer's scorecard.
| Measure | Tests | Formula | Level |
|---|---|---|---|
| On-time delivery, OTD | Date | Deliveries inside the window ÷ deliveries due | Order or line |
| Case fill, or unit fill | Quantity | Units shipped ÷ units ordered | Units |
| Line fill | Quantity | Lines shipped complete ÷ lines ordered | Line |
| Order fill | Quantity | Orders shipped complete ÷ orders | Order |
| OTIF | Both | Orders, or lines, on time and in full ÷ orders, or lines, due | Order or line |
Four orders, ten lines. Window: requested date, plus or minus one day. In full: 100 percent of the line.
| Order | Line | Ordered | Shipped | On time? | In full? |
|---|---|---|---|---|---|
| 1 | 1 | 100 | 100 | Yes | Yes |
| 1 | 2 | 50 | 50 | Yes | Yes |
| 1 | 3 | 20 | 20 | Yes | Yes |
| 2 | 1 | 200 | 180 | Yes | No |
| 2 | 2 | 40 | 40 | Yes | Yes |
| 3 | 1 | 60 | 60 | No, 3 days late | Yes |
| 3 | 2 | 30 | 30 | No, 3 days late | Yes |
| 3 | 3 | 10 | 10 | No, 3 days late | Yes |
| 4 | 1 | 400 | 400 | Yes | Yes |
| 4 | 2 | 90 | 0 | Yes, for what shipped | No |
On-time delivery, order level = 3 of 4 = 75%. Line level = 7 of 10 = 70%. Case fill = 890 ÷ 1,000 = 89% Line fill = 8 of 10 = 80% Order fill = 2 of 4 = 50% OTIF, line level = lines both on time and in full = 5 of 10 = 50% OTIF, order level = 1 of 4 = 25%
Only order 1 passes both tests on every line. One set of deliveries produces seven figures from 25 to 89 percent. A supplier reporting case fill and a retailer scoring order-level OTIF are both right, and they are 64 points apart.
OTIF is the intersection. If late orders and short orders were the same orders, OTIF would equal the lower of on-time and fill. When they are different orders, as here, where order 3 is late and orders 2 and 4 are short, the misses add up. The identity makes it visible:
Orders = OTIF + late only + short only + both = 1 + 1 + 2 + 0 = 4
The split across late only, short only and both is the first cut of cause: late only points at transport and scheduling; short only at stock and picking; both at a stock-out that delayed the order and still left it incomplete.
| Customer type | Usually scores | Date basis | Charges? |
|---|---|---|---|
| Large grocery and mass retail | OTIF, order or case level | Requested, tight window | Yes, a percentage of the cost of goods on failing cases |
| Mid-size retail | OTIF or fill plus on-time, separately | Requested | Sometimes |
| Foodservice operators | Order fill, with substitutions rules | Delivery day | Rarely; they change supplier |
| Industrial and trade | On-time to promise; line fill | Promised | Rarely |
| OEM, just-in-time | On-time to the hour; 100 percent fill assumed | Scheduled slot | Yes, line-down charges |
Find out which, per customer, and compute that version beside your own. The OTIF benchmark page covers the required levels, and the Excel guide computes line and order level with the customer's tolerances in separate cells.
The best version reported. Case fill at 97 on the monthly pack while the customer's order-level OTIF is 82 and the chargebacks arrive.
Promised date as the basis. The promise was moved when the stock was short; every order is on time to a date the customer never wanted.
Substitutions counted as filled. The line shipped, with a different product. Whether that is in full is the customer's rule, not the supplier's.
Backorders counted twice. The short line ships later as a new delivery, which is on time to its own new date, and lifts on-time performance.
On-time tests the date, fill tests the quantity, OTIF tests both and is always the lowest. Report all of them at stated levels, lead with the one each customer scores, and use the late-only, short-only, both split to decide who owns the fix. For the arithmetic by hand, see OTIF on ten purchase order lines. Covirage computes every version from the order and delivery files, per customer, with the identity checked.
Because the misses do not overlap. If 8 percent of orders are late and a different 7 percent are short, on-time is 92, order fill is 93, and OTIF is 85. The closer OTIF is to the lower of the two, the more the same orders are failing both tests, which usually means a stock problem causing both the delay and the short.
All three, because they answer different questions. Case fill tells you about volume service and is the highest. Line fill tells you about range availability. Order fill tells you what the customer experienced, since one short line makes the order incomplete. If the customer scores you on one, lead with that one.
It depends on the customer. A retailer with a promotion needs the date. A contractor needs the whole order to start the job and would rather wait a day. The cause split, late only, short only, both, tells you which kind of failure you mostly have; the customer tells you which kind hurts them.