On-time delivery (OTD) is one of contract manufacturing's most common measurements, but we rarely talk about how difficult it is to measure. Here's a guide to defining and measuring OTD.
Basic Definition
OTD rarely refers to a specific date; it usually refers to a range of dates. In most companies, OTD refers to a range of dates defined as X days before (early) and Y days after (late) the due date. A typical OTD window is 5 days early, 0 days late (can be expressed as -5+0). If, for example, an item is due June 1, it would be considered on time if it arrives on any day between May 27 and June 1. This concept of an OTD window is embedded in almost all modern ERP/MRP systems.
Definition Factors
The two main factors that influence the OTD window are production line requirements and cash flow. If a particular delivery is expensive, it may be planned for delivery very close to the production need date so it cannot be late. Since it is expensive it cannot be early, so a tight window is appropriate. If an item is inexpensive and planned in bulk (for example floor stock of screws) it can arrive within a very wide window and be considered on time. The key is that whatever your definition is, it is clearly understood by your supply base.
Commodity Class - It's common for companies to class components by value, commonly referred to as A items (high cost) B items (medium cost), or C items (low cost). If your system has the flexibility, you can set the OTD window by commodity class. Common windows are -5+0 for A items, -5+1 for B items, and -10+5 for C items. It is critical to note that these windows must be coordinated with the planning lead times in ERP. You cannot have +5 day window for C items if the ERP is driving C items to arrive on the day of need.
Pull Systems - Pull systems includes Kanban, Min/max, bin replacements, etc. Most associated with variants of Just In Time (JIT) programs, pull systems are any approach that involves delivery triggered by an event. By contrast, in traditional push systems delivery is triggered by a PO date. We do not advocate the use of OTD for pull systems; the appropriate performance measurement approach is stock outs. However, due to limitations in ERP environments companies often must try to apply an OTD metric to pull systems. In this situation, we suggest using blanket POs for annual, quarterly, monthly, or even weekly quantities. Set your early window to equal the performance period of the blanket. As a starting point, you might choose to issue blanket PO line items for one month requirements. In this case, set the early window to -31 and the late window to 0. If you are experiencing experiencing stock outs, you will need to set the blanket window to one week. If no stock outs, you can loosen the blanket to one quarter.
Measurement Factors
Defining OTD is the starting point, but over time measuring OTD is the real challenge. Reliable OTD statistics that everyone, internal and external to your organization, can count on are the only stats worth tracking. Here are some of the challenges.
Working days vs. calendar days – This is one of the most common misunderstandings between customer and supplier. Does 5 days early include weekends? Holidays? It doesn't really matter which method you choose, as long as the suppliers understand. That said, we recommend calendar: it is intuitive, easy to understand, and most widely adopted. Whichever you choose, make sure your ERP is aligned. Often the working day or calendar day decision will effect the entire ERP planning module.
Ship date vs. dock date - Another basic misunderstanding is, does the date used to measure OTD refer to the date the item is shipped, or the date the item is received? It's best practice to use the dock date, the date the item is received. This the most intuitive and the most widely used. The supplier should be accountable to consider transit times when determining what ship date they will need to make in order to supply the material within the OTD window.
Promise date vs. required date - A promise date is the date the supplier committed to meet. A required date is the delivery date needed by the customer. The promise date is best practice and by far the most widely used. It is rarely fair to judge a supplier to a requirement they have not agreed to. Using required dates for OTD measurement is appropriate when certain agreements are in place, for example vendor stocking programs or guaranteed lead times that were pre-negotiated.
Original promise date vs. revised promise date - The promise date (the original delivery date) will often be altered at some point during the order process by either the client or manufacturer, and will then be substituted with a revised promise date. Standard calculation of OTD takes both original and revised promised date into account when measuring success; however, this is not always the case, and customers may only utilize original delivery date when calculating OTD.
Percent of line item vs percent of quantity ordered - A line item is a single line in a purchase order that specifies a product. In order for delivery to be considered on time, each separate item must be closed, meaning the order is filled. The percent of quantity ordered is calculated in parts per million (PPM) and is generally considered filled when the customer has received 99 percent of the requested order.
Purchaser vs. supplier data - Often a supplier's measurement of OTD varies from the customer's. All of the factors above can contribute to such a difference. The best thing that can be done is to periodically reconcile each other's measurements. We suggest this should be done quarterly, even more frequently if large differences persist.
Through careful consideration of the above factors you can ensure an accurate OTD calculation that is a helpful metric for both you and your suppliers.