Delivery Metrics: On-Time, In-Full & Fill Rate
A customer sees one order as a simple promise: “Will it reach when you said, and will everything be inside?” The warehouse sees the same order as inventory availability, picking accuracy, dispatch cut-off, carrier capacity and last-mile execution - any one weak link can break the promise.
That is why delivery metrics are not “logistics numbers.” They are the sharpest test of whether a supply chain can convert demand into reliable customer experience.
- On-time delivery measures whether orders arrive within the promised delivery window.
- In-full measures whether every confirmed item and quantity was delivered.
- OTIF is stricter than both: the same order must be on time and complete.
- Fill rate measures demand served immediately from available stock, usually at unit or line level.
- A company can have high fill rate but poor OTIF if it ships most units but misses time windows or order completeness.
- Track delivery metrics with reason codes - stock-out, pick error, carrier delay, address issue, capacity cut-off - or the KPI will not improve.
- The interview trap: calculating on-time and in-full separately, then assuming the customer experience is good.
Big Picture: Six Measures Behind the Customer Promise
Delivery performance has two sides: speed reliability and completeness reliability. A customer does not mentally separate them. A late complete order and an on-time incomplete order are both broken promises.
Core Explanation: How the Metrics Fit Together
On-time delivery asks: did the order arrive within the promised date or time slot? It is usually measured at order level, not unit level, because the customer experiences the whole order as one promise.
In-full asks: did the customer receive every confirmed line and quantity? If a grocery order had 20 confirmed items and 19 arrived, the order is not in-full even if the missing item is small.
OTIF - on-time in-full - asks both questions on the same order. This is why OTIF is usually lower than on-time rate or in-full rate alone. It is the stricter, more customer-realistic metric.
Fill rate is different. It measures how much demand can be served immediately from available stock. It is more of an inventory availability metric than a last-mile delivery metric. If fill rate is weak, OTIF will eventually suffer because warehouses and stores cannot ship what they do not have. If you need the inventory-policy logic behind this, revise setting inventory policy for a multi-product business.
The One Numerical Example You Must Be Able to Do
Assume a company received 1,000 orders in a week. Out of these, 920 were delivered within the promised window, 900 were delivered with all confirmed items, and 850 satisfied both conditions. Customers demanded 10,000 units, and the company immediately shipped 9,600 units from available stock.
The learning: fill rate can look healthy while OTIF looks weak. That usually means the problem is not only stock availability - it may be picking errors, split shipments, carrier misses, wrong cut-offs or poor order promising.
Definitions: Say These Clearly in One Breath
- On-time delivery: Percent of orders delivered within the promised delivery date or time window.
- In-full delivery: Percent of orders delivered with every confirmed line item and quantity supplied.
- OTIF: Percent of orders delivered both on time and in full, using the same order denominator.
- Fill rate: Percent of demand fulfilled immediately from available stock, before backorders or lost sales.
- Perfect order rate: Percent of orders completed without time, quantity, damage, documentation or billing defects.
How to Diagnose a Delivery Metric Problem
Do not stop at “OTIF is low.” Break the problem by failure type, because each failure has a different owner.
For recurring stock-linked misses, a pull system such as Kanban and pull-based replenishment can help convert actual consumption into timely replenishment signals.
Case Study: BigBasket and Slot-Based Grocery Delivery
BigBasket shows why delivery metrics must combine inventory accuracy, picking discipline and last-mile slot reliability in one customer promise.
Situation. Online grocery is unforgiving. Customers often order milk, vegetables, staples and household items for a specific day’s consumption. A late delivery hurts convenience; a missing item can force a second purchase trip. So the real promise is not “we shipped most units” - it is “your chosen basket arrives in your chosen slot.”
The move. BigBasket’s slot-based model makes capacity visible before the promise is made. The primary driver is controlled order promising: customers choose available delivery slots instead of receiving a vague delivery estimate. Supporting drivers include fulfilment centre or dark-store inventory visibility, batch picking, cold-chain handling for perishables, substitution rules and route planning.
The lesson. The business cannot manage on-time delivery in isolation. A perfect slot still disappoints if key SKUs are missing; a perfectly picked basket still disappoints if it misses the dinner-time window. This is exactly why OTIF is a better customer-experience metric than separate on-time and fill-rate numbers.
So what: BigBasket’s delivery reliability comes chiefly from promising only what operations can fulfil in a slot, supported by inventory visibility, picking discipline and routing control. That is the answer interviewers like - primary driver plus supporting drivers, not a one-cause story.
How AI Changes Delivery Metrics
AI does not replace delivery metrics. It makes them more predictive. Instead of discovering tomorrow that OTIF fell, companies can flag at-risk orders before the cut-off and intervene while the order can still be saved.
The best student workflow: load a company description, a delivery-metrics dashboard and the job description into NotebookLM, then ask it to generate “five likely interview questions on OTIF, fill rate and delivery exceptions.” For the analytics angle, connect this topic with using AI for inventory optimisation and replenishment.
Interview Relevance
“A retailer has 96% fill rate but only 84% OTIF. What could be going wrong, and how would you improve it?”
Use one sentence that signals maturity: “I would not celebrate 96% fill rate until I know whether the missing 4% is concentrated in high-value SKUs, key customers or promised orders.”
Common Mistake
The mistake is treating on-time, in-full and fill rate as interchangeable. It costs candidates because they miss the customer-level truth: an order can score well on one metric and still fail the promise. Fix: calculate OTIF on the same order denominator, then use fill rate and reason codes to diagnose why it failed.