Push, Pull...or Flow
In my most recent email newsletter (send me an email to subscribe) I recounted an online discussion I recently took part in regarding push vs. pull manufacturing. Everyone agreed that pull was better than push but there was lots of discussion about what, exactly, were the definitions of each.
Here's an excerpt from that newsletter.
A "push" approach is where things are made before a customer (or the next step in the process) wants it. A "pull" approach is where things aren't made until a customer (or the next step in the process) asks for it. In general, then, agile thinking is that pull is better than push.
But there's a lot of ground between these simple definitions and practice. Let's say a company has an order for 10,000 widgets to be delivered 1000 each month for the next ten months. Let's also say that the company makes all 10,000 widgets at once. Is that push or pull? The company didn't make the widgets before a customer ordered them, so it's pull, right? On the other hand, the company made 9000 of the widgets before the company wanted delivery of them so it's push. And push is bad.
Let's say that our widget company goes to strictly pull; no product made before its time. And in the third month, the customer says, "My business is going through the roof. I need 3500 of those widgets this month." Or the supplier says, "I know I promised that raw material for your widgets this month but we've had huge tooling problems. I can get it all to you next month." At that point, our widget company will be wishing it had stuck with it's old push approach.
The tough part of all this that our widget manufacturer has competing demands: figuring out how to maximize customer service while keeping costs as low as it can. It can maximize customer service by having inventory in the warehouse ready to ship (push). It can keep costs low by not having any inventory (pull). But it's tough to do both.
Rather than push or pull, I prefer to talk about flow. Does material flow from our receiving dock (raw materials) to the customer's receiving dock (finished goods) in a smooth, steady way? This doesn't mean that it never stops (sits in inventory) but it does mean that the flow through all steps of the process is steady, predictable, and, yes, as fast as possible without causing harm to the overall system or its outputs.
I'll have more to say in future posts.
Here's an excerpt from that newsletter.
A "push" approach is where things are made before a customer (or the next step in the process) wants it. A "pull" approach is where things aren't made until a customer (or the next step in the process) asks for it. In general, then, agile thinking is that pull is better than push.
But there's a lot of ground between these simple definitions and practice. Let's say a company has an order for 10,000 widgets to be delivered 1000 each month for the next ten months. Let's also say that the company makes all 10,000 widgets at once. Is that push or pull? The company didn't make the widgets before a customer ordered them, so it's pull, right? On the other hand, the company made 9000 of the widgets before the company wanted delivery of them so it's push. And push is bad.
Let's say that our widget company goes to strictly pull; no product made before its time. And in the third month, the customer says, "My business is going through the roof. I need 3500 of those widgets this month." Or the supplier says, "I know I promised that raw material for your widgets this month but we've had huge tooling problems. I can get it all to you next month." At that point, our widget company will be wishing it had stuck with it's old push approach.
The tough part of all this that our widget manufacturer has competing demands: figuring out how to maximize customer service while keeping costs as low as it can. It can maximize customer service by having inventory in the warehouse ready to ship (push). It can keep costs low by not having any inventory (pull). But it's tough to do both.
Rather than push or pull, I prefer to talk about flow. Does material flow from our receiving dock (raw materials) to the customer's receiving dock (finished goods) in a smooth, steady way? This doesn't mean that it never stops (sits in inventory) but it does mean that the flow through all steps of the process is steady, predictable, and, yes, as fast as possible without causing harm to the overall system or its outputs.
I'll have more to say in future posts.
Comments