Demand Management Process for Steady Material Supply

During one of my stints working for a client in the US, the team was setting up the design for the best practices in demand management. As most of us are aware, the performance of any company reflects on how the demand management process works. Demand is the point of entry for the rest of the operation. If inventory is building, lead times are increasing and delivery dates are being missed, then the demand management is the first place that the folks have to understand and control. The strategies used for demand management is dependent on industries – as retail industries look for solution to capture the POS data to get a real-time semblance of the demand pattern. Asset intensive industries are dependent on how well they manage the S&OP planning process to determine the market addressable based on the rough cut capacity planning.

During the design planning session a team consisting of sales, planning, marketing and IT were gathered in a room. As we have probably realized that demand, itself, is an ambiguous term. People talk about – forecast (shipment forecast, sales forecast, booking forecast, unconstraint forecast), sales orders, inter company transfer orders, procurement forecast and purchase orders. It took the team a while to differentiate between the various types of forecast and what has to be the prime driver to manage the sales and operations planning process.

The other challenge for companies is the capture the information for all these types of demand. The most common element of the forecast is the shipment information that is historically archived. In a commodity business this is a good start to create a forecast during the annual planning process. The forecast based on the shipment is constraint based, clearly indicating what the company has shipped in the past, and is a good measure of generating a forecast for the future demand. The sales and marketing teams can use this forecast as the starting point to modify based on market intelligence reports and changes in the customer buying pattern. The modified forecast becomes the actual sales forecast that the business planners have to corroborate and conduct the rough cut capacity planning.

The following discussion will be mostly about the forecasting process. The bookings forecast, as conducted by sales department, on the other hand used in many companies are a way to hedge the sales forecast errors. Bookings forecast can be higher than the sales forecast if the sales think that a new order may show up before the end of the month or at the beginning of the next month. However, the bookings is not evil that has to be eliminated but should be controlled by the planning team. In the commodity business, where the demand is cyclical and seasonal, it is essential to level the production based on the demand variability. The planner has to determine the build up of the inventory or booking of capacity to meet future demand.

There are several best practices in managing the forecast, but based on my experience the key elements that a forecaster has to grasp is the reliability and the repeatability of the forecast. Assuming that the forecast is reliable with a high degree of forecast accuracy, which can be measured, based on the company business and processes and proper risk mitigation strategies should be applied. From a repeatability point of view, if the forecast is fairly level, it is easy to plan the supply to meet the demand and control the level of inventory in WIP as well as in the finished goods. When the demand is seasonal e.g. spikes in the demand at the end of a period, it should be treated carefully to ensure a steady supply without causing a situation of a stock-out. If the supply is flexible and reliable then the process can be run with minimum inventory. The risk for Just-In-Time increases as the rate of consumption increases and the supply has to keep up with the consumption. Best practices suggest that a steady supply is the best way to operate which will mitigate the risk of stock outs. Inventory planning, based on the repeatability of the forecast, plays a critical part to determine the level of inventory to maintain, which will keep the cost of inventory down as well as maintain high customer service levels. The repeatability of the forecast can help the inventory planner to size the finished goods inventory in the warehouse and the distribution centers thus promoting steady supply and high customer service levels. If there are any feedbacks or questions, please let me know.

Stability of Supply Chains

Recently I had some interesting discussions with executives from numerous companies who are at different stages of moving forward in their Lean journey. The business drivers for most of these clients are how to reduce cost and the manufacturing lead times.

After meeting with several clients and doing a bit of research it seems that if there is one metric that companies should focus on is the lead time reduction. Lead time is the core metric that influences the health of the operations. People familiar with “Little’s Law” would quickly grasp that there is a direct correlation between Lead time and WIP in the system. As the WIP is steadily increased in the system the Lead time does not change. This indicates that some level of WIP is good for the supply chain, and the operations can keep running with the same level of operational performance. But at the inflexion point, any increase in WIP will adversely affect the Lead time. What it indicates that some level of WIP is definitely good but in excess it is the cause of the increase in the cost of operation. Excess WIP, requires more handling cost, loss of material, increased material movement, more quality problems due to handling and an overall increase in the complex decision making for scheduling.

Where is your operation with respect to the inflexion point? A typical company never reaches the efficiency as dictated by Little’s Law. It is a theoretical limit and gives perspective for companies to compare their operations to the theoretical limit. The typical reasons for a company to deviate from the limits are a) Increase in product mix, b) batching requirement are each operation, c) volatility of the supply and demand, d) setups and e) not producing parts/components based on customer requirement.

In this highly variable environment, what can companies do to run effectively? Notice – I mention effective and not efficient. Companies can run their operations efficiently by producing material at the resource capacity. This does not mean that the products they are producing are based on customer needs. The end customer can be the next operation, distribution centers, channel customers or end customers. Running effective operations indicate good business practice and produce the right mix of material that can run their operations efficiently as well as meet the customer demand.

The main crux of the manufacturing effectiveness is the 3rd principle of Lean as defined by “James Womach” in his book “Lean Thinking,” which is making the material flow through the system. These concepts are not new, we see this in real life with fluids dynamics. The only way to reduce the residence time of the fluid is to make it flow, thus reducing the stagnation points as defined as “Sinks”. The same thought process can be applied in Manufacturing – we want to avoid “Sink” or material stagnation points. I am not preaching that the sinks should not be there, it is necessary to have these sinks to manage product portfolio, variability in supply and demand, or the manufacturing scheduling constraints e.g. calendar mismatch, rate of production, required mix of production etc. It is however a good process to determine the size of these “sinks” and manage them so that we do not loose control.

This gets us to a new point – how do you compensate or measure the operations managers to run “effective” operations. Do you measure them based on the deviation from the predicted “sinks” or inventory buffer target or do you measure them on resource utilization?

Efficiency at the Cost of Flexibility

During my discussions with clients in the past the focus was primarily efficiency rather than flexibility. I have to admit most of the clients that I worked with belonged to an asset intensive industry where millions of dollars were invested to install the asset. The focus of the operation was to keep these expensive asset running to reduce the unit cost (based on utilization) but sometimes at the cost of loosing control over unwanted inventory buildup.

We had a unique experience the other day at McDonald. Dallas was unusually cold during the holidays and we planned to take the kids to the neighborhood McDonalds to burn some calories (sounds like an oxymoron) – in the indoor play area. We had a bunch of 4 boys who wanted to play tag. The play area at a McDonalds is a colorful group of slides, walk-in pipes, climbing nets and steps combined together to form a pleasing area to play. The two 11 year olds formed a tag team and the two 6 year olds formed the other team.

At first thought it seemed that it was a slam dunk where the 11 year olds would win easily. They were stronger, faster and bigger. They could climb up easily, take bigger steps and could overpower their brethrens to surrender easily. What we did not realize that the walk-in pipes had a limited diameter. It was easy for the 6 year olds to run through them while the 11 year old had a tougher time and their movement was at best very sluggish. It was soon proven that the 6 year olds won the tag as a team – while being physically weaker of the two teams.

It is a good example where it was evident that flexibility is a key enabler to speed up the operation. Manufacturing philosophies have changed since Frederick Taylor promulgated the art of improving efficiency through mass production in his research on “Scientific Management”. Henry Ford and Andrew Carnegie were early adopters who took advantage of the philosophies building standardized massive installation to build the high efficacy machines. The role of manufacturing was based on push production and was thrived on low mix and less global competition. Ford could keep on building efficiently their black Model-T’s in dedicated lines, and still be able to sell them in the market. But in today’s market the role of customers have taken the center stage. Customers dictate the product requirement and the attributes. Customers have become accustomed to better customer service and have consistently challenged the producers to shorten their lead times. The discussion to shorten the lead times in a high product mix environment is itself a topic of discussion.

Flexibility is a virtue that manufacturers have learnt to adopt in the new developing global economy. There are several ways to improve flexibility in operations. One of the ways is installing several smaller machines instead of a single large machine. It might cost more during investment but will increase the flexibility. If the organization is based on a smaller machine environment, some of the machines will always be up while the others are being maintained. Flexibility aids producing several different kinds of products simultaneously through it manufacturing lines. Flexibility also improves if you are planning to your rated capacity based on past reliability records of the equipments. Plan based on the effective capacity instead of the rated capacity. The rule of thumb used by planners is to plan around 85% of capacity leaving some room to accommodate unreliability of machines and labor, machine setups due to product mix changes and customer order expedites.

Based on the observation, I think I would rather have the two 6 year olds in my team rather than the two eleven year olds. They are smaller and more agile in the given environment. One could argue that if the walk-in pipes bigger the results would be different – well “that” is the market demand – we have to learn to be flexible and adjust to it.

Promising in Finer Time Buckets

Depending on the industry vertical and the “lead time” of manufacturing or services, the promising of orders are in days, weeks or months. The aerospace industry will promise the next airplane delivery in months, the steel industry in weeks, the auto parts industry in days and the IT project implementation in months. The question that the VP of operations often poses is how to promise better in finer time buckets.

During one of my stints in Asia, the CIO who came from a manufacturing background wanted a solution/strategy to promise his customers in days instead of the usual way of operating which was weeks. The benefits of promising and delivering in days was very simple – it would take away a week of inventory from the supply chain which would reduce the lead times and also leave extra cash of the table that can be put to more productive use rather than being tied in inventory.

Though the reasons for promising delivery is finer buckets is very simple – but the task of achieving it is pretty complex. During further investigation on why the company was promising in weekly buckets we found that there were basically 3 main reasons. The first reason was that the extra buffer of time was to cover the unreliability of the operations, the second reason was that each operation in the manufacturing chain was scheduled to maximize the utilization of the resource thus sometimes resulting in the production of unwanted material or producing material before hand and the third reason was that they wanted to ship in large quantities aggregating to reduce the shipping cost.

The first task at hand is to investigate the root causes of reliability. The reason can be varied, ranging from quality problems, machine uptime, poor planning and demand variability. The key to eliminate reliability is to systematically eliminate all the root causes. The other way to manage reliability issues is by creating a buffer of safety stock that can sustain the variability. Although safety stock is additional capital tied up in inventory – it does enable creating a flow of material through the manufacturing chain. The key to any manufacturer is to decide on the safety stock level and not just any stock – but a predetermined amount for a mix of stock. Safety stock is definitely not the panacea to eliminate reliability, but it does help bringing stability and consistency to the production. As further actions are taken to eliminate reliability, the safety stock can be reduced and the capital released to be used elsewhere.

The second factor is the fascination about utilization, of machines and people, by operations manager. The relentless focus on utilization reduces the focus on flexibility. No operation is perfect in the imperfect world. Creating a plan for 100% utilization increases the risk for unplanned disruptions. A small hiccup in the operation can lead to a ripple effect which can last a significant amount of time with loss of customer service, escalation of orders, complex decision making for available capacity and overall stress on the people and machines. All of this results in nervousness in the supply chain and results in the increase in inventory. In the imperfect world the advantages of running an operation at 85-90% capacity far supersedes the benefits of running at 100% capacity. It provides flexibility to take care of unforeseen interruptions and allows the goal of providing excellent customer services at a lower cost. The concept of process design and line design by industrial engineers take into account the variability in demand before setting up the manufacturing process. The premise of the design is to enable flow of material through the line thus reducing the latency within the production line. However, the rate of change in the business climate has increased several folds over the years. The original typewriter had a lifecycle of over 30 years – while the current technologies have a much shorter lifecycle. Business processes and manufacturing lines have to be reconfigured to align to the new business climate every so often to keep itself tunes to the new dance. Running an operation at less than 100% capacity provides the ability to react and accommodate incremental changes in the business while maintaining excellent customer service.

The third factor as explained was the consolidation of shipment into large batches. This is again a counterintuitive approach to improve efficiency. Larger batches do help reduce the cost per unit. But there are inherent costs associated with accumulating goods to reduce shipping costs. Material have a tendency of not adhering to the planned timeline and a delay to a portion of the shipments will delay the entire shipment. Keeping goods in the warehouse, while waiting for the batch to be complete, increases the risk of goods being stolen, lost or damaged. Increase in goods as resident increases overall material handling costs – tracking and tracing, inventory management data entries, not finding the right material and others. A smaller shipment can keep the customer operating, while waiting for a large batch before shipping increases the risk for the customer. An unplanned event during the shipment can derail the entire shipment while if shipped in smaller batches can reduce the overall risk for the customer.

The benefits for promising in finer time buckets are great, but require discipline in bringing stability, consistency, repeatability and sustainability to manufacturing. Operations Manager should systematically address and eliminate the root causes of reliability, run the operation smoothly and below 100% capacity and ship in smaller batches.

I would love to hear from you about your experiences.

Running Operations Lean

The story exemplified in Ely Goldratt in his famous book “The Goal” about a team of “Scouts” on a camping trip was an excellent example exhibiting the role of a bottleneck and its impact on the overall progress of the team. The slowest member in the chain controls the pace of the progress and the key focus of TOC (Theory of Constraint) is to manage and improve the slowest member or the bottleneck. TOC helps manufacturing companies to identify the bottleneck and focus resources to elevate the constraint. Lean experts look at the same example a different way. Lean fundamentals explore and understand the supply chain. Lean would have taken the same team of “Scouts” and determined the rate of each member – analyzed the “takt” time and balanced the carrying load for each member of the team so that the entire team moves at a uniform pace. Lean approach to manufacturing is ensuring that the processes are balanced and run per the “takt” time as determined for the customer demand.

Lean principles will enable the team to move at a uniform pace with minimum amount of inventory (distance) between the members. Lean takes a holistic approach to design the effective supply chain (line balance) and recognizes the nuances in the supply chain constraints to create a uniform flow across every unit in the chain. TOC on the other hand is a more focused approach to address the bottleneck. TOC is a good solution for factories which have less control or foresight into the demand product mix and are continually re-planning to determine the optimal way to schedule the demand to meet delivery performance. Usually this model, the variability in production causes reliability issues for the people, process, machine and materials in order to satisfy the four requirements of an efficient and effective supply chain is cost, quality, safety and delivery performance.

A good analogy of a Lean and TOC practice is the human body. Every person has a different metabolism of absorbing the calorie intake. In a Lean model, the person has knowledge of its metabolism and creates a corresponding diet plan. The diet plan, if maintained, will ensure that the body remains lean. The diet becomes a habit and all corresponding activities of ensuring the right diet is available and regular health checkup are planned easily. A regular exercise schedule is laid out which ensures that the body stays in shape. The exercise schedule is based on the variability in the calories intake. If the variability is high then a regular strenuous exercise is laid out – which buffers that body against the variability. Variability in the diet is taken into account while planning the exercise routine. TOC model is useful when there is a lack of regimental schedule and the body has to react to the changes in the diet. The emphasis is not on the calorie intake but on the process of burning the calories. If on a certain day the calorie intake is high, the body meter indicates that the exercise schedule has to be strenuous. If TOC determines that exercise is the bottleneck, it will plan for extra time on the stair-master, to burn the extra calories. Lean and TOC both address the metabolism and try to keep the body lean. Lean dictates that the body metabolism has to be understood and set up all the activities, including diet and exercise plan, to meet the body’s metabolism requirement, while TOC will quickly address the variability by addressing the bottleneck and focus on making it more efficient. Both the processes require continual evaluations to reset the regiment. Although it may look better on paper that one could only focus on the bottleneck rather than take a holistic approach, it causes uncertainty and irregularity in the schedule which can become cost prohibitive in the long run.

The human body is an embodiment of perfect manufacturing environment. It is a learning system which processes massive amounts of data, provides visibility, makes decisions and reacts to variability.

Love to hear if you agree or disagree.