Main content

John J. Civerolo

See if you can spot this manufacturing company's problem: The first thing every morning there is a two-hour "shoot out" meeting. The purpose of this meeting is to try to figure out what needs to be made that day to meet the customer's needs. They are trying to balance the Demand = Supply equation. The company is trying to figure out what needs to be done in each work center, then generate a schedule that is followed on the production floor.

Unfortunately, soon after the morning meeting, they have "Earl--the expediter extraordinaire" running through the plant hourly changing the schedule as fast as he can. This is the way many manufacturing companies operate today. In many manufacturing companies there are a number of Earls, some in executive positions. However, the objective of this article is to show that companies do not have to operate in this inefficient and non-value-adding manner.

In the tough competitive times ahead, it will be important to be proactive, instead of reactive. To meet customers' needs, the company must adjust the factory to the customer--not adjust the customer to the factory. But, it is also very critical to do this without having lot of work-in-process (excessive queues), expediting, firefighting, backlogs behind every work center, past due manufacturing schedules, backorders, missed & past due customer orders, capacity overloads & underloads, material shortages, people standing around not knowing what to do next, or people having too much to do and still not knowing what to do next.

Five Simple and Key Questions

Let's look at what the production floor wants to know daily:

  • Which one should be made next?
  • When should the next one be made?
  • How many to make?
  • Where should we make them?
  • How to build them?

Not real complicated, right? The first question can be answered by a schedule. Unfortunately, a schedule is often incorrectly thought to be synonymous with a computer report. A schedule could exist without a computer or a printout. In fact, we know one manufacturing company where the schedule is handwritten daily.

Another company communicates the schedule by a schedule board. One company communicates the schedule by Kanban signals. In another, the schedule is communicated by a simple color code. Here is how the color code system works:

The schedule calls for producing 100 different individual items, all to be done this week. Hopefully, 20 items per day will be done. The color of the week is green. A colored sticker is put on the customer order packet information that travels with each item. 

At the end of the week, if there are still five green jobs with green stickers on the production floor, what does that tell you? Only 95% of the schedule was met. Five are past due and missed their due date.

The color for the next week is blue. If, at the end of the next week, we walk the production line and find both green and blue colors on the line, it says we are one and two weeks past due on some items. If we find four colors (green, blue, yellow, and red) at the end of any week--well, that says the production floor is in deep trouble. Simple, right? The production floor finds this method easy to manage and understand. Plus, it is very visual and out-of-balance conditions are easily spotted.

Whether you use colored stickers, schedule boards, Kanban signals, handwritten schedules or a computer printout, the critical factor is having a schedule that is valid, realistic and achievable.

The second question the production floor wanted answered was: When should the next one be made? This can be answered by one of two dispatching rules:

Push rule - Make the next item immediately after making the last item.

Pull rule - Make the next item only when authorized by the downstream (internal customer) workcenter.

There are many ways to authorize (signal) the upstream (internal supplier) workcenter to start producing:

  • A card
  • Empty tubs, boxes, bins, racks, spools, carts, pallets, pegs, or squares on the floor
  • Colored golf balls
  • Computer signal (sometimes called electronic Kanban)

Is push bad and pull good? Pull bad and push good? Sounds like the great beer debate: Tastes great--less filling. Who cares--it is only beer! Push or pull are only simple alternative dispatching rules.

The third question is answered by the lot size, order quantity, replenishment quantity, batch, campaign, etc., that matches the customer needs with the manufacturing capability. The objective should be to make the exact quantity the customer needs, not more or less.

The fourth question is answered by the line, work cell, work center, department, machine, operation, etc., where the product is to be run productively and cost effectively.

The fifth question is answered by having high quality and on time documentation (drawings, specifications, procedures, routings/processes/sequence of events, method sheets, etc.) in production’s hands when they are to use this information to build the product.

Let's explore the push and pull dispatching techniques to better understand how to apply them.

Push or Pull: What is the Difference?

The following is a brief explanation of how the push and pull dispatching rules work:

Push logic. Suppose there are 100 blue items and 100 red items waiting to be built in WIP inventory (queue) in operation #30. When the 100 blue items are completed, they are "pushed" to the downstream customer. Which ones do they make next? The red items. When do they make them? Right then! How did they know which ones to make first? The schedule told them.

Pull logic. There are 100 blue items and 100 red items in WIP inventory (queue) sitting in front of the operation #30. Labor has already been added. A good way to look at this inventory in queue is that it is stored capacity and material. The downstream customer pulls the 100 blue items and sends a signal to workcenter #30. If the logic is to make more of what is taken, then the workcenter makes more blue items. This is how Kanban answers what to make next. When do they make the next ones? Only after they have received the signal.

There have been many discussions about the advantages and disadvantages of each of the rules. Some companies feel the push rule has the following advantages:

  • Keeps people busy and they can make the efficiency standard.
  • Utilizes the equipment and keeps it running.
  • Absorbs the overhead cost.

The disadvantages are:

  • The downstream work center may not want, have a need for, or have a place to store the items that were pushed to them. Didn't the upstream work center care about their internal customer's situation? No! In many cases, the challenge is to bury the internal customer with work because of individual or work center performance measurements. Reminds you of separate non-linked operations, instead of a team working in a synchronized flow environment.
  • Could result in increased work-in-process inventory.
  • Could waste critical resources (dollars, people, capacity, material & time) on inventory that is sitting in queue.

Some companies feel the pull rule has the following advantages:

  • Links operations. The internal customers/suppliers are in synch because the upstream produces only what the downstream needs and when they need it.
  • Inventory levels are forced to remain at a given level. The inventory levels are determined by a business strategy that matches the external customer’s needs and expectations and keeps a balanced and synchronized manufacturing flow.
  • Value is added only to items when they are needed.


The disadvantages are:

  • People are not always busy producing inventory and may be idle at times. They are temporarily not producing.
  • Equipment may be temporarily shut down.
  • Overhead may not be absorbed, creating an accounting variance.

Many times people feel it is the dispatching rules (push or pull) that either cause or solve problems. Simply not true! The problem of excess inventory or reduction of inventory is not in the dispatching rules.

The root cause of the problem is the way the production floor is directed and managed, the existing financial measurements, and the massive behavioral change required to do things differently (thinking outside the traditional box) on the production floor. Many companies don't address the root causes of these problems, thus causing problems with excess inventory, past due customer orders & manufacturing schedules, and long cycle (lead) time. Other companies do address and solve the root causes of problems. That is the key--not the dispatching rules.

There are two major issues not commonly understood:

1. Balanced Capacity

Too often the needs of the customer (internal & external) are not in balance with the capabilities of the production floor. Overloads (more requirements than have consistently been demonstrated by production) or underloads (not enough requirements) exist. An objective of meeting the customer requested date 100% of the time is to have supply equal demand. Capacity (people & equipment) must be available and have consistently been demonstrated. Out of balance capacity will over or under load the production floor, regardless of the material replenishment tool or technique. There must be a capacity balance between the workcenters, line, cells, departments, operations, machines, etc., to ensure a synchronous flow. If there is an imbalance, then it results in more inventory, longer cycle times, poor customer service, wasted resources, etc. This is why proactive capacity planning with visibility into the future must exist.

2. Quality Schedules

Let's go back to the first question: Which one should be made next? A schedule could communicate this information. Unfortunately, many companies don't have a high quality, valid and realistic schedule. [See Figure 1]


Figure 1

Part #     Due Date     Today: 5/21/04

Blue         7/25/03

Black     12/05/03

Red         1/02/04

Yellow     1/15/04

White       2/02/04

Purple     2/16/04


The root cause is that the schedule in Figure 1 doesn't pass the "sanity check"--it is invalid, unrealistic, unachievable and confusing! Everything is past due. Which ones should the production floor work next? The latest date? No way! I was told by a production operator that jobs have been canceled, but remain on the schedule & on the floor. So, which ones are really needed? The production people will probably have to wait for the expediter, expediting meeting, or “hot list” to communicate the real needs. I recently walked a production floor and saw jobs that had been in WIP for months. Scarce resources had been applied to these jobs. Now, that’s waste!

One of the first things a manufacturing company must do is to have a realistic, valid, and high quality schedule. [See Figure 2]


Figure 2

Part #     Due Date     Today: 5/21/04

Blue         5/21/04

Black        5/21/04

Red           5/24/04

Yellow        5/25/04

White         5/26/04

Purple        5/27/04

See the difference in the two schedules? Which one would you rather work with on the production floor?

Once the schedule is valid, the question of which one to make next can be answered.

There are three ways to communicate the schedule and answer the two production floor questions. They are:

Pure Kanban. Sometimes called Demand Pull. This technique answers two questions: Which one should be made next? and When should the next one be made?

Hybrid. A schedule answers the question of which one should be made next. The pull rule answers when to make the next one.

The schedule could be communicated by a dispatch list. This answers the question on which one to make next. The pull dispatching rule answers the questions of when to make the next one.

The focus of this article will be on the Pure Kanban (Demand Pull) and Hybrid techniques. Let's discuss each one to see what is needed to effectively use them successfully.

Pure Kanban

Here is a very simple example of how Pure Kanban works: There has been a designated level of inventory (with value added) sitting in the work center. Note: Someone in the company decided the level of inventory based on some business or part-by-part strategy (based on inventory turns/spins, dollars & units), not the Kanban technique or a magical number. Whatever the internal customer (downstream work center) takes is replaced. If a red one (or a full box of them) is taken, another red one is produced. If a blue one is taken, another blue one is produced, etc. Very simple, right? Absolutely! Whatever is consumed is replaced when consumed. The pure Kanban technique is forecasting that more demand will soon be required. If this suddenly reminds you of the old reorder point, two-bin system, or how your wife/husband helps generate the grocery list--well, it should. Because that's what it is. This technique has been around for years. Do you know what the original Kanban signal was? It was a milk bottle. Leave the exact number of empty bottles out and the milkman would replace that number with full ones. Therefore, I suggest that the milkman originally came up with the Demand Pull (Kanban) idea.

The Pure Kanban concept is easily understood. But, what isn't always as easily understood is that a number of prerequisites should be in place to help make this communication and material replenishment technique effective. They are:

1. Pull Rule is in Effect.

The pull rule says to make the next one when authorized, through a signal, by the internal customer (downstream work center). The advantage is that the internal supplier is in synch with the internal customer. The internal supplier only makes items when needed, not sooner. Both outstanding benefits! If there is no authorization, (signal) from the internal customer, what does the internal supplier do? Temporarily STOPS producing! What do the people do? Hopefully, something productive rather than being sent home without pay or just continuing to produce inventory. These productive activities could include: solving quality problems, reducing non-value-adding activities, cross-training, reducing setup/changeover time, being involved in action teams, thinking outside the traditional box, brainstorming improvements, doing root cause analyses on existing problems, etc.

A person in manufacturing recently asked, "Are companies really suggesting that people stop producing?" See the problem? This person's focus is probably on the old traditional efficiency and utilization measurements & meeting the standards (always keep people and equipment busy). Temporarily stopping production could cause a major problem in many companies if the required behavioral changes haven't taken place on the production floor and/or at the executive level. Accounting systems will also have to be changed. There also needs to be a different set of performance measurements.

A company recently implemented the Pure Kanban technique. Hours of Why education and How training took place. Many specific company examples were reviewed by everyone on the production floor. But, the first day Pure Kanban was implemented "stuff happened”. The downstream workcenter had an equipment breakdown. The signals being used were squares on the floor. Four squares, with inventory, in front of the upstream workcenter were authorized. By 10:00 A.M. on the day shift, the four squares were full. What should the upstream workcenter do? Temporarily STOP producing! Well, they didn't follow the Pull rule and kept producing inventory by the tons. They were asked why they didn't stop.

Their answer: I would have had to stop at 10:00 A.M.! They repeated this answer over and over. See the problem? They really didn't believe it was all right to temporarily stop producing inventory and shut down the equipment. So, it is critical that the pull rule be understood and adhered to. This is very important. Stopping was forbidden in the past, now it is okay as a smart way to control inventory, keep cycle times realistic, and manage the business & the critical company resources.

One key to success is limiting, controlling, and managing the queue (buffer) inventory, and not allowing more inventory to be produced unless the downstream workcenter needs it. If the inventory is managed, controlled and limited, it doesn’t really matter which dispatching rule is used. If the inventory was limited to two boxes, pieces, spools, pallets, cartons, carts, tubs, etc., does it matter whether you push or pull? Absolutely not! So, the key to success is limiting, controlling and managing the inventory, not pushing or pulling.

In some companies, they become controlled by the Pull rule. “Follow it or die”. They feel it is blasphemous to not follow the Pull rule. But, the reality of business and not missing customer orders or requiring additional cost to strictly follow a rule may not pass the sanity check. This is why in many companies they allow an “override” to the Pull rule to keep a balanced flow of material and productivity. It may require the temporary use of the “Push” rule to come through the crisis. These “override” rules need to be approved and authorized by Operations management. Once the crisis is passed or the root cause of the problem causing the crisis is eliminated, the Pull rule is again in effect.

2. Repetitive Usage.

Whatever is taken is immediately replaced. The assumption is more is needed soon. Demand Pull (Kanban) is forecasting that more demand will soon be required. This may not always be true. This is only a valid assumption if the item to be replaced has repetitive usage. If the item is infrequently used, a one time special, or has seasonal or fluctuating demand, this assumption could lead to a pile of excess & obsolete inventory. Pure Kanban can't be used in these environments. The Hybrid Kanban technique can be used to overcome this limitation.

3. Very Few Items.

If there are many different items, then a real problem could develop. A given amount of floor space and a designated inventory quantity would be required for each item. The results could be a very expensive way to manage the business--taking up a lot of floor space with inventory. Again, the Hybrid technique can be used to overcome this limitation.

4. Small Order Quantities (Lot Sizes).

Demand Pull (Pure Kanban) signals when and what to make--not how many. That is a separate question. How many to make is an order quantity (lot size) question. Making large order quantities could waste a lot of capacity (people & equipment) and material. Typically, if larger order quantities are not properly managed, it could result in an increase in inventory, longer manufacturing cycle times, misuse of many critical capacity (people & equipment) & material resources, quality problems, and more non-value-adding activities.

Beware! Before just running out and reducing the order quantities, there are a number of obstacles that must first be resolved. They are:

  • Long setup/changeover times.
  • Extra paperwork, transactions for inventory receipts & issues, work orders and/or purchase orders.
  • Frequency of material handling.
  • Transportation costs.
  • Financial performance measurements:
  • Efficiency.
  • Utilization.
  • Inventory treated as an asset versus a liability.
  • Absorb overhead with direct labor.
  • Unit cost mentality--absorbing the cost of the setups/changeovers over the run quantity to drive down the individual piece costs.
  • More inspections.
  • Non-synchronized/balanced flow environment. Disjointed processes and material flow and out-of-balanced capacity.
  • Old attitudes and mindset where the "bigger is better" philosophy has to be changed to a "smaller is better" philosophy.
  • Reducing lot sizes before overcoming these obstacles could drive up costs, increase inventory, cause poor customer service, increase cycle time, increase non-value-adding activities, and drive down acceptance of a valid new idea.

5. Very Short Cycle (Lead) Times.

When does the internal customer expect their product? When they turn around to get it! They assume their needs will be met on time. The expect that material is “worry free”. If not, then production delays occur. If it takes a long time for the internal supplier to produce the item, then the Kanban signal must be sent before the last item or container is used up.

The net results are more inventory and all the disadvantages of excess inventory. The root causes of long cycle times should be attacked with a vengeance so lot sizes can be significantly reduced.

6. Reliable Process.

If the inventory buffer is reduced, then there is little or no protection when "stuff happens." When the internal customer goes to use the material, how good does is have to be? Excellent! Why? Because, if it isn't, many delays occur. This means that the process must be in control and must produce quality (defect free) material each and every time.

If the quality of the material is bad, the following could happen:

  • Rejects and scrap
  • Rework
  • Delays
  • Equipment down
  • Longer cycle times
  • More cost
  • More expediting
  • More schedule changes
  • Missed customer delivery
  • More queue (WIP)

A potential solution to the above problems is to trigger the Kanban (Pull) signal earlier, which adds more inventory--again defeating the objective of smaller inventories and reduced costs. But! It may be required to meet the internal and external customer’s needs.

The key to success is to address the root causes of the poor quality of material and remove them. Then, less inventory and “buffer” is needed to protect against variability in the process and poor product quality.

7. Excellent Material Planning.

When the work centers turn to the stockroom, what do they expect? “Worry free” material! The right material on hand, in the right quantity, at the right location, and on time. How does this occur? Magically? Of course not!

Kanban--without the right material on hand, in the right quantity, on time to meet the demands--means more of what was done in the past: Expediting, hot lists, mismatched sets of parts & material, excess inventory, daily shoot out meetings, etc. Getting the right material doesn't happen by accident.

A good Sales and Operations Planning process, valid master schedule, a high quality bill of material, excellent inventory accuracy, good material planning linked to the execution process, external supplier visibility of the requirements, and good on time performance are critical to the success of Pure Kanban or meeting any schedule.

8. Excellent Capacity (People and Equipment) Planning.

Here is a problem. See if you can tell what caused it. There are two authorized containers, one green and one purple, between the upstream work center (internal supplier) and the downstream work center (internal customer). The downstream work center gets a customer order for a green product and pulls the green container. The upstream work center is signaled that another green is needed. They start to produce another green. The downstream work center completes the green product and ships it. They receive another customer order for a purple product. Now, a purple is authorized. But, the upstream work center is still working on the first green one. The downstream work center finishes making the purple product and ships it. Another customer order is received for a green product. The downstream work center looks for the green container. Nothing there!

Unfortunately, the upstream work center is still working on the first green authorization. Now, the upstream work center has a green, purple, and another green authorization. What's the problem? The capacity between the work centers is out of balance! The work centers are not synchronized. The downstream capabilities are faster than the upstream. The upstream work center is not capable of satisfying the downstream internal customer’s demands. There are a number of solutions that could be employed to solve the problem:

  • Slow down the downstream work center or speed up the upstream work center.
  • Move some people from the downstream to the upstream work center. What does this assume? That the people are cross-trained and the equipment is available for them to use.
  • Add more equipment -- could be a long-range versus a short-range solution.
  • Increase demonstrated capacity and productivity by reducing the non-value-adding activities: setups/changeovers, scrap, rework, schedule changes, material shortages, wasted time, missing tooling, equipment downtime, people absent, etc.
  • Work overtime.
  • Add an extra shift. Again, this assumes trained people can be quickly added.
  • Increase the inventory buffers by adding more green or purple product. The problem: When do we produce the extra inventory and how much should be produced? Could be very costly, and again, defeats the purpose of small order quantities and small inventories.

The problem could have been prevented if, before rapid fire signals piled up, the upstream work center or the external supplier had visibility of capacity needs through excellent proactive capacity planning versus reactive capacity planning after the trouble has caused production & customer problems. Remember, the milkman could find himself a little short of cows if he pulled into your driveway and found 2,000 empty milk bottles on the front lawn without any advance warning! The lead time to install more cows and increase capacity could be a major problem. The same things will happen in a manufacturing company or at a supplier without proactive capacity planning.

For Demand Pull (Kanban) to work successfully, visibility of future capacity is needed before the Kanban signals show up.

Keys to Success

Now, let's ask two important questions: If we have repetitive items, excellent material planning, excellent capacity planning, small lot sizes, very short cycle times, short setup/changeover time, synchronized & balanced flow, reliable processes, and we pushed the material instead of pulling--would we be successful? Of course! See, the key is short manufacturing lead time, small order quantities, reduction of setup & changeover time, removing the root causes, synchronizing & balancing the flow, etc. The key is not the specific dispatching technique used.

The second question is: Does it matter if we push or pull, as long as the queue (buffer inventory) is managed, limited, controlled, kept small & spinning, and there is a high quality schedule? The answer is NO!


There is a problem for many manufacturing companies if they don't have repetitive usage and very few items as a result of the "standardizing" effort. They need to know how to answer the question, "Which one should be made next?" Instead of replacing exactly what is consumed, a schedule communicates which one to make next. Demand Pull (Kanban) then signals only when to make the next one, not which one to make. Which one to make is answered by a valid and quality schedule. So, companies without repetitive usage and very few items use the Hybrid method to overcome these two obstacles to using Pure Kanban. It is called Hybrid because it is a combination of Kanban signals and a schedule--a little of both.


The Pure Kanban and Hybrid techniques both have the following advantages:

  • Simple--both techniques are easily understood.
  • Excellent material planning and movement techniques. They both do an excellent job in signaling when to make more.
  • Problems surface. It is easy to spot a problem since inventories (queues) are managed, controlled, and limited. Inventory is not used to cover up or hide problems.
  • Visual capacity control. Since the inventories (queues) are controlled, limited and managed, an out-of-balance capacity situation is easily seen.  
  • Demand Pull or Hybrid techniques do not automatically reduce queue. As problems are solved and inventory (queue) is brought under control & reduced, then it is possible to systematically shrink the amount of inventory that is needed for buffer or protection against “stuff happening”.
  • Ordering material is moved out on the production floor and is done by production floor people, which eliminates an overhead activity.
  • Helps reduce cycle time. As the non-value-adding activities are attacked with a vengeance and queues (inventories) are reduced, then cycle time can be significantly reduced.
  • Less paperwork, fewer inventory transactions & moves, and more visual controls.


Many companies get outstanding benefits with Demand Pull (Pure Kanban). What is the success attributed to? Is it the Pull technique? Was it the cards, squares, or other signals? What was the real key to success? Success is achieved by changing the way the business has been traditionally managed in the past. Successful companies focus on reducing the non-value-adding activities: setups/changeovers, rework, scrap, material shortages, excessive paperwork, poorly maintained equipment, quality problems, etc. They work on getting valid & high quality schedules, managing & shrinking the inventories (queues), making sure demonstrated capacity is equal to required capacity, having a synchronous & balanced flow, and changing the traditional performance measurements such as efficiency & utilization. They stop keeping people and equipment busy to absorb overhead or meet the efficiency standard. They stop producing inventory to show a paper profit. They also redesign the product to standardize and reduce the variety of items to meet customer needs, they re-lay out the manufacturing process using cellular manufacturing to cut down on material movement & setup/changeover time, they simplify & streamline the complex processes, etc.

Here is what all of this will mean, expressed by one manufacturing company President, "The '90s will be very exciting and extremely challenging. The challenge will be to become marketplace-driven and have a strong customer focus, while being very flexible & agile in the factory and by our supplier base. Flexibility and agility must be achieved without increasing the cost of running the business, having excess inventory, and having long cycle times. No longer can our manufacturing company be capacity and/or material constrained. Material and capacity must be “worry free”. The factory will have to adjust to the customer. We can no longer afford to adjust the customer to the factory and remain competitive." If properly implemented, the Demand Pull and Hybrid techniques can help meet these business objectives.

Too often, the Demand Pull technique was blindly implemented without the proper understanding and without ensuring the prerequisites needed to make Demand Pull work effectively were in place. The results were disappointing and sometimes disastrous. Companies lost confidence in a powerful and extremely effective technique.

If your manufacturing company is using or is thinking of using the Demand Pull (Pure Kanban) or Hybrid techniques, are the prerequisites for success in place? If not, Beware -- your company may be heading for a failure or disaster that could affect customer satisfaction and bottom line performance. Before implementing the technique, ensure your team is working on implementing the prerequisites for success.

Does this mean a company doesn't start until all the prerequisites are in place? Absolutely not! Start out identifying the obstacles preventing shorter manufacturing cycle times, smaller lot sizes, reliable processes, excellent capacity/material planning, product variety, etc. Once they are identified, determine the root causes & prioritize them. Inventory buffers may be required until the root causes are removed. Focusing on reducing or eliminating the root causes one at a time is important. As the root causes are removed, reduce the inventory buffers.

The final result will be happier customers with on time delivery of a quality product and a lower cost. This is the challenge for the '90s.



If you have specific questions about this article or want to discuss it with the author, call Chris Gray at 1 603 778-9211. 

The Partners for Excellence specialize in helping companies set up comprehensive measurement programs and improving overall resource management performance.  Contact us at 1 603 528-0840 or emailThis email address is being protected from spambots. You need JavaScript enabled to view it..