Quoting reliable lead times in production without guessing on gut feel

Short answer. You can only quote a reliable lead time once capacity, stock and open orders sit in one view. You calculate the lead time on real capacity: free hours on the bottleneck machine, available raw materials and what is already scheduled. Not on gut feel or a fixed standard term.

Most lead times in manufacturing are a guess. Someone looks at the order book, senses how busy it is and says four weeks. Sometimes that is right, often it is not. And when it is wrong, you pay it back in rush orders, overtime and a customer who hesitates the next time. I build systems that turn that gut feel into a calculation, so the date you quote is also the date you hit.

Why lead times based on gut feel structurally go wrong

The problem is not the planner. It is the fragmentation. Capacity sits on a planning board or in someone's head, stock in another system, open orders in the ERP or in Excel. Nobody sees those three things at the same time the moment the customer calls. So you quote a date based on one of the three, usually the order book, and ignore the other two.

That works fine for a long time on the happy path. It goes wrong on the exceptions: the order that needs a raw material still on its way in, the week your best welder is on holiday, the rush job that got squeezed in. Those exceptions are exactly what decides whether your lead time is reliable.

Calculate lead time on real capacity, not on a fixed term

A reliable lead time comes out of three questions you have to be able to answer at the same time. How many free hours do I have on the bottleneck machine or the bottleneck department in the coming weeks. Will I have the raw materials and semi-finished parts in on time. What is already scheduled and cannot be shifted. Only once those three sit in one view can you name a date that is based on something.

In practice your lead time often depends on one or two critical capacities, the bottleneck. Once you know where it sits, the calculation becomes manageable. You reserve capacity on that bottleneck resource, add the throughput time and account for the lead time of the materials you still have to purchase. That is no longer a gut feel, it is a sum.

First review the process, then start calculating

Before I build anything, I look at how the planning runs now. Many steps in a manufacturing company simply grew that way. The planner manually pulls a list every morning, copies it into a personal Excel, calls the shop floor for the current status and quotes lead times based on that. Part of that work is duplicate or redundant once the data sits in one place.

Sometimes an entire step turns out to be unnecessary, for example retyping orders between systems. Sometimes it turns out you do not need a new planning system but a connection between what you already have. That is a deliberate step, not an excuse to do nothing. The calculation has to happen. The only question is which route leads there most cleverly.

Standard package or custom build: where the line is

For make-to-order manufacturers, good planning packages exist. An APS such as Limis Planner calculates finite capacity and connects to ERP systems, and Exact offers production planning with material and capacity needs in context. For many companies that is exactly enough, and I say so when it is.

But a standard package covers the happy path. It calculates neatly as long as your production fits the vendor's model. Your exceptions rarely fit fully into that: the deviating operation sequence, the way you reserve capacity for quotes, the cost allocation per order you need for your post-calculation. Those are exactly the points where the lead time becomes reliable or unreliable.

That is where I build the solution that does fit your situation. Sometimes that is a dashboard that brings capacity, stock and orders from your existing systems together into one lead-time view. Sometimes an AI workflow that automatically checks incoming requests against capacity and proposes a substantiated date. Sometimes a connection that feeds your APS the shop-floor data it cannot see on its own. The moat sits in the custom work on the exceptions, not in the package around it.