Participative Sales and Operations Planning a new method of planning

This article describes a participative approach to the process of Sales and Operations Planning including balancing load and capacity, managing batch sizes and lead times, and using simple spreadsheets to plan jobs rather than sophisticated Advanced Planning and Scheduling Systems APS. It covers, Master Production Scheduling (MPS) and Rough Cut Capacity Planning (RCCP). Two companion articles accompany this article relating to "Capacity Management" and "Demand Management".

This paper was reproduced in its original form in the House Magazine of the UK Institute of Operations Management "Control" in December 2000. Our latest thinking includes the coordination of sales planning, product introduction, product changes and a control system to manage the process. This makes a more complete definition of the requirement to be "Participative Development, Sales & Operations Management", which we now teach in our training courses and an overview of which can be found in Capacity Management.

Links to related training and further reading on left

What is your capacity? The computer software suppliers will tell you it is finite. Is it? In what circumstances is it finite? What does finite mean? How can it be influenced to improve it? What are the current influences that adversely affect it? What happens as your demand approaches your capacity? What is the best way of balancing load and capacity?

The finite capacity generally referred to is one of the three constraints described in Capacity Management namely "Hard Ceilings". Hard ceilings are where the capacity is extremely difficult to flex to accommodate increased volume. I have encountered two types of hard ceiling:

A major piece of capital equipment which runs at a fixed rate such as a mixing bowl, ball mill, or heat treatment, where process times are fixed, or production line where the track rate is fixed. In this case it is financially prohibitive to buy another to increase capacity. All you can do is run it flat out, 24 hours a day, 7 days a week, avoid breakdowns (frequency and duration) and quality problems, and ensure that it is always working to customer needs and not building stock, which wastes utilisation.
A job, requiring a scarce skill that is difficult to train such as, hand-crafted articles, or indirect workers such as toolmaker, or maintenance operatives. There is a limit to how much overtime can be worked to meet demand and the training programme to reach basic skills is protracted.

In both cases it is difficult to increase output above a given level and sub-contracting is not practical for quality reasons, or lack of availability of suitable sources.

Soft ceilings can be flexed by buying additional inexpensive plant, recruiting unskilled or semiskilled staff, or subcontracting, or of course short-term overtime.

The essential differences therefore between the soft and hard capacity ceiling are the cost of flexing and lead-time to flex ("capacity action response time"). (See Capacity Management.) "Finite" therefore refers to this.

(We discuss the other two other types of capacity constraint in our S04 Strategic Capacity Management training course, namely "hard floors" (high fixed costs) and "hard walls" (bandwidth or mix sensitive capacity), which may also be relevant to medium term S&OP.)

CAPACITY LIMITS

Next I will deal with the problem of demand approaching your capacity limits.

Filling Capacity & Lead-time Management

As the order book fills and capacity is allocated to these orders, observed lead-times increase. Conversely as the order book contracts, observed lead-times reduce because the available capacity is idle. (See Capacity Management (Development, Sales & Operations Management). Failure to reflect this simple fact in planning has the following effects:

Effect of Arrears / Back-orders

Load vs. capacity over time

A typical output plan when viewed as a capacity model on key (bottleneck) resources

(The output plan is sometimes referred to as "operations plan" / "Master Production Schedule", or the diagram also describes a typical order book load in make-to-order situations)

The customer is still expecting their order on the date originally promised. If overdue orders / arrears are not rescheduled in conjunction with the customer they create a short-term overload of equal due dates, which induce "thrashing" (see the capacity hockey stick below). In scheduling, a scheduling tool simply needs some guidance on priority before it can produce any schedule. "Finite" capacity planning does not help here unless the conflicts are communicated to customers. (The effect of Finite scheduling will be to simply delay some orders already promised earlier. The effect of prioritising however can result in extreme delays on jobs with a low priority. In one case where we did this, a 9 month delay.) (See Previous technique T037: FIFO (How FIFO reduces the onset of thrashing))

The capacity hockey stick

When lead-times are not managed, overloads will occur. What happens is illustrated in the following diagram. This shows that as capacity is reached, the operations system starts to "thrash".

"Thrashing" is the problem of the system keeping itself busy re-planning rather than producing, which effectively reduces capacity. What happens in this situation is that the system becomes "interrupt driven" similar to mainframe computers where this problem was first documented. This analogy is a good one because interrupt is an accurate way of describing the customer panic interrupts which disrupt the process to satisfy today's crisis.

The graph shows a gradual increase in lead-time as load increases up to the point where "thrashing" occurs and where lead-times go through the roof! Exponential effect of load on lead-time

Thrashing results in the familiar problems of:

Arrears / Backlogs
Dissatisfied customers
Everything urgent / expedited
Everything a bottleneck
Increased transport
Shortage lists High stock / WIP
Priorities confused
Split batches / reduced productivity

There are a number of circumstances where thrashing can be self-induced, with the obvious adverse effects. These include, a number of operational malpractices, quality problems, plant and equipment unreliability, motivation of the workforce, skills management, and computer based scheduling which will not be further pursued here. However improvements in these before the crisis occurs will defer the onset of thrashing.

It is important to understand the underlying reasons in defining the solution. We discuss possible causes and solutions in more detail in the 3rd article of the series on reducing lead-times, Best Practice No. B015: Lead-time reduction 3: Arrears and Backlogs. Because these symptoms can be confusing, it is essential that a diagnosis of the problem is undertaken as outlined in the article "Diagnosing Manufacturing Control Problems".

Affect of lead-time ratio (demand vs. supply)

If supply lead-time is close to demand lead-time there is little scope to schedule and any delay will cause overloads and thrashing. This is where we have found the "Scottie" approach useful, if customers will accept longer lead-times. Captain Kirk on the Star Ship "Enterprise" in the middle of a battle asks Scottie when the engine will be fixed. Scottie replies: "It is extremely difficult. It will take a day sir", (knowing full well that it will take about 2 hours). He delivers in 4 hours, the battle is won and Scottie is a hero.

Product Life Cycles

Unfortunately lead-time can be a self fulfilling prophesy. If you say it will take a week, and plan to do that, even though there is only a day's worth of work, it will still appear in a week's time!

When a new product is launched the lead-times are in fact longer than they will be later when, the learning curve will have been climbed, supplier relationships established, snags removed from the design, etc. Quite often lead-times are loaded to planning systems such as MRP systems at this point and are frequently forgotten. At the end of a product's life with reduced volumes, lead-times can increase due to supplies being more difficult to obtain. (It becomes a special rather than a fast mover.)

The overall message here is that you must:

Constantly target lead-time (See Agile Manufacturing)
Actively maintain lead-times
Reflect current lead-times in promises to customers.

ACTIVELY MANAGING BATCH SIZES BASED ON AVAILABLE CAPACITY

Using Pareto analysis, increase batch sizes of the low volume value "C" items, and decrease the batch sizes of the fast movers. If you have not done this before you can expect simultaneous increases in customer delivery performance and reduced stock levels. The benefits will vary according to whether you currently have higher than necessary stocks or lower than required customer service levels. There are some detailed considerations here both in terms of the feasibility of this in some environments and implementation, but this is the underlying principle of Level Scheduling.

Forget EBQ / EOQ. Use ATQ (Available Time Quantity) to calculate batch sizes calculated as follows:

Time to complete the whole mix over a planning cycle

=	The operational time to do all of the batches
	+
	The time to change over between all of the batches

Provided the time available is greater than this you can reduce batch sizes further. (And don't forget to keep working on changeover times.)

This formula can be applied to manufactured parts, where the in-house capacity constraints are considered, or purchased parts where buying office, goods inwards, logistics traffic and supplier constraints are considered. (Changeover time in the buying office equates to number of orders handled per period). (A full explanation is given in Previous Malpractice M005: EBQ the worst way to set batch and order sizes) (We also show you how to model the effect of batch size in SSC03 Advanced Forecasting & Inventory Modelling.)

Volume Mix Exchange Curve

The following diagram illustrates the capacity trade off between mix and volume.

Trade off between volume and mix

If only one variety is being produced output can be more than if several varieties are being produced due to the influence of changeovers. The parallel graphs are due to step changes in capacity bought about by moving the ceiling. In this case the company had the option to introduce double line manning, and extra shifts in response to highly seasonal demand. I.e. to move their firm but not hard ceiling. Provided that demand is below the exchange curve it is a feasible plan. Using exchange curves can simplify the assessment of particular loads. If your current demand is close to the curve it is time to take action.

MASTER PRODUCTION SCHEDULING

This brings me to Master Production Scheduling where I have found that a simple approach to "Rough Cut Capacity Planning" based on bottleneck capacities and a "Participative Master Production Scheduling" process works in most circumstances. You need an active Master Production Scheduling system to avoid accepting orders with unrealistic lead-times.

Rough Cut Capacity Planning

First it is worth defining the environments where this process will be inadequate.

I have discovered one circumstance where only finite capacity planning of the type afforded by APS systems will actually do the job and the characteristics of that environment are as follows:

It was complex with, a high product mix (hundreds of products), a number of alternate methods (work centres) of processing available, but each work centre had different processing times.
It had very precise and tightly defined process times for each alternative work centre.
The product structure (bills of material) and methods of manufacture (routings) were locked and subject to strict change control.
Process reliability and quality were excellent (almost perfect).

The situation I have defined is a mathematically programmable but complex scheduling problem, but not requiring statistical probability calculations arising from uncertainty. My first reaction to this situation however was to consider segmenting the products and processes into more manageable chunks, to simplify the problem. (See Organisational Design)

Any uncertainty caused by absenteeism, quality, process reliability, etc. complicates the application of scheduling tools, and makes their resulting schedules unstable. Since the APS school claims that the ability to re-plan is a virtue, I need to stop and explain. Flexibility in planning provided that is mirrored in production and the supply chain is fine. If it is not, it is a method of building unwanted work in progress. (See Previous technique T037: FIFO (How FIFO reduces the onset of thrashing))

The are two simple methods of Rough Cut Capacity Planning:

The first and simplest is "Takt Time" is described in Previous Technique T021: "Takt Time".
The second is a spreadsheet based technique which is described below.

A spreadsheet will suffice for up to several hundred products over several periods, although the maintenance of it becomes a problem at this point. In particular aligning the product load profile (see later) maintained within the spreadsheet, with a downloaded schedule from your sales system becomes difficult unless the load profile can be maintained within the sales system as well. Another situation requiring a more sophisticated approach is where there is a high product mix with routings that are significantly different for each product. In this situation loads can appear on resources at different times due to the position of the resource in the routings. (The lead-time offset). In both cases it is worth considering a more sophisticated tool, but our advice is to try doing it this way first, before you decide that your problem is too complex to manage this way. You will learn from the process and be able to justify the additional expenditure!

In most cases consideration of multiple constraints is actually a waste of time, since in our experience bottlenecks generally do not move regularly, except in the unique situation I have just described.

Having now dismissed these circumstance as unusual, in our experience, I will now describe the method of capacity planning and scheduling which works for most manufacturing businesses and avoids the need to spend lots of money on computer systems and their expensive support.

As shown below, starting with the bottlenecks create a spreadsheet describing the capacity consumed on the bottleneck resource by each product. (A product load profile).

Across the spreadsheet maintain the required customer quantities. Calculate load hours (or other specified units of load) by multiplying the (hours) per product by the quantity in the period to create (hours) per period columns.

With simple products the quantity alone may be enough to assess the feasibility of the plan.

Product	Load Profile on bottleneck	Period 1	Period 2	Period 3 Etc.	Period 1 Load	Period 2 Load	Period 3 Load etc.
A	Hrs/product	Qty	Qty	Qty	Hours	Hours	Hours
B	Hrs/product	Qty	Qty	Qty	Hours	Hours	Hours
etc.	Hrs/product	Qty	Qty	Qty	Hours	Hours	Hours
Total load		Total	Total	Total	Total	Total	Total
Demonstrated Capacity		Qty	Qty	Qty	Hours	Hours	Hours
Under / overload		Qty	Qty	Qty	Hours	Hours	Hours

Use the spreadsheet to calculate the load for each product in each period and total.

Compare this to demonstrated capacity and highlight the differences. Demonstrated capacity is simply, previously achieved output, where it is expected that this can be repeated, or (only if in a step change situation) the planned output.

This model may be refined for multiple constraints by creating a spreadsheet that contains a multidimensional-constraint, load profile.

We supply the working spreadsheet template (illustrated) with our training courses: S04, M05, SSC08

RCCP Model

In this case a product load profile is maintained for each product for up to nine resources. Each resource load is calculated by period by multiplying the hours per product by the quantity required per period to give the load per resource per period. In the illustration above the plan is valued by period & the load per period compared to the capacity available, with over / under loads highlighted.

To define the load profile, ask shop floor supervisors. Use the routing times as the last resort and only then if the shop floor supervisor agrees with them.

Participative Master Production Scheduling

The traditional approach of Master Production Scheduling requires a Master Production Scheduler who calculates the schedule and presents his results to production to make. It is a non-participative process particularly if a computer based scheduling tool is used in the calculations. Our approach used in many situations now is to employ more participation from the stakeholders. It requires a planning meeting following the first cut schedule, (based on the initial spreadsheet output).

The purpose of the meeting is to:

To debate options, agree, commit to and own the plan for the future period, taking into account the present state of play and upcoming events.
To review and change capacity, batch sizes and lead-times as necessary.
To continuously improve the planning process and the capacity model.

The meeting is formal with minutes and after the settling down period should last no longer than an hour. This can only be achieved by proper preparation. It is through this meeting that capacity is actively managed. It is important therefore that all the stakeholders attend and that they have the authority to act. Stakeholders from sales and production will always be present. It is also common for Purchasing to attend if supply lead-times or capacities are critical. In some circumstances critical suppliers should attend. But I have known engineering attend in capital equipment manufacture, and where new product introduction or engineering change is significant. I have known accounts attend if expenditure control is important, and HR if recruitment and training are important.

Power is removed from the master production scheduler, and given to the meeting. If the first cut schedule is a poor fit to customer requirements or capability, or in unusual circumstances, it may require a second round of calculations and meeting to agree a schedule. In practice this is quite rare, following the early implementation phase.

Seasonal Businesses

The agenda and review frequency needs to be considered carefully, around the time of the ramp up and ramp down periods.

Implementation

The consideration of the main bottleneck is the first step on the road to refining the capacity model. Therefore it is better to start with a simple model quickly than to try to get it completely right first time. Demonstrated capacity is measured at the end of each period. The meeting is then used to identify refinements needed to both the model and the planning process.

Sales Forecasts

Sales forecasting is purely the responsibility and domain of sales and marketing. However the master scheduling process ultimate aim is to meet the orders or forecast and use the available resources effectively.

Conclusion

Sales and Operations Planning is a business process, not a set of computer tools. It is not a job done by an individual (the master scheduler), although the process requires administrative support. In some unusual cases, computer modelling, based on APS type tools, is useful, but is rarely essential. Sales & Operations Planning can be reactive to sales needs, in which case Participative Master Production Scheduling (operations planning) is appropriate in manufacturing environments.

The components of the process include:

Understanding load and capacity and managing them
Understanding product life cycles, and lead-times, and managing them
Avoiding "thrashing"
Managing batch sizes
Rough Cut Capacity Planning supported by appropriate computer tools if necessary, using as simple as possible capacity models

It is controlled through an active Participative Master Production Scheduling Process. It does not include sales forecasting.

Conversely Sales & Operations Planning is a more all-embracing process which can be used in addition to drive a sales process, driven by a number of factors including bottom up sales forecasts, for which there is sales accountability.

Participative Development Sales & Operations Management (DS&OM) takes that a stage further into a complete capacity management control system.

_____________________________________________

Bookmarks for this topic above:

Our full range of training

Relevant Training / Workshops

Expert Systems / Tools

Relevant Further Reading

Top

To discuss your consulting or training needs with one of our independent consultants or trainers please Contact Us

Whilst great care has been taken to provide relevant, accurate, practical, advice based on our considerable process design and development experience, this will almost certainly require interpretation into the context of your unique business. Please be careful in doing so and if in doubt seek expert advice. We would welcome your feedback!

SM Thacker & Associates

Independent Best Practice Training & Consultancy

Highlights of our full range of training courses / Workshops:

Bookmarks for this topic below:

Relevant Training Course / In-house Workshop Highlights:

Expert Systems / Tools:

Relevant Further Reading: The following further articles were mentioned in this paper:

Participative Sales & Operations Management (S&OP)