It occurred to me that the productivity improvement debate could benefit from an alternative perspective
Whilst everyone might be using the same term, it is not necessarily so that we have the same meaning , agree on what it is that we as seeking to apply productivity to and whether the the terms ‘productivity’ and’ cost reduction’ are as interchangeable as we would like to believe
What is productivity?
Before I do anything, let me declare a useful and non controversial meaning of productivity; a measure of output per unit of input.
Inputs include such items as labor, capital and time and their associated costs, while outputs can include revenue, GDP or a myriad of other variables.
Similarly, my definition of a system is that it is a set of activities that are working together (interdependent) in order to achieve a common purpose. Clearly this definition relies on the system having a common and higher purpose that unites those (local) activities within that (global) system together and creates the need for interdependence. Implicit in this definition is that whether there is a conflict between the purpose of the global system and the purpose of the activity, that the former takes precedence over the latter.
What is important, is that if we are serious about measuring the productivity of a system we must ensure that there is alignment between the system’s purpose and those numerators (outputs) and denominators (inputs) of the productivity ratio. For example, if the system’s purpose is to deliver profit,the numerator needs to some form of revenue (gross, net, contribution margin) and the denominator needs to be costs (operating expenses) associated with that revenue over an agreed time frame.
Of course the calculation can become problematic (but solvable) when there are multiple inputs and/or outputs but the principle remains the same.
Improvements in productivity can occur by either increasing outputs relative to the same level of inputs or a decrease in inputs whilst keeping outputs constant or a combination thereof.
Are we all referring to the same system?
Global systems (like a ‘for profit’ organisation) have sub processes and sub-sub processes within them and so it is somewhat important that we are clear about what we mean when it comes to defining the system that is subject to productivity measurement/improvement. This is because it is quite possible for a sub process to generate more output relative to its input without increasing the global organisations financial performance. For example,
– the maintenance department could maintain more units of equipment per working hour and actually result in a net increase in the overall cost of maintenance
– Supply could cut the cost of inventory by refusing to hold slow moving stock but this could result in lost sales in the case of extended equipment breakdown caused by having to wait for the delivery of those same spare parts.
– a business could cut costs by reducing overtime but simultaneously cause revenue to be lost because that same overtime was to be used to process a backlog of customer orders
Local activities can only be truly productive when the activities that they perform result in an improvement in the global outcome. Nothing more, nothing less. The most effective form of measuring a local activities impact on the global system is to measure them in terms of
a) Not doing the things that they should have been doing (Throughput dollar days – TDD)
c) Doing things that they shouldn’t be doing (Inventory dollar days – IDD)
So if a local productivity improvement doesn’t result in making money for the global system, then it should not be classified as being a productivity improvement initiative (let alone allowed to continue).
Does the system’s constraint matter?
Because systems are composed of interdependent activities it means that there is typically one or very few of these activities that limit the throughput of the entire system. It therefore logically follows that if we can improve the throughput rate of the systems ‘constraint’ we are in fact increasing the throughput rate for the entire system.
Productivity from a systems perspective therefore measures the rate at which work flows through the system constraint ie. Throughput per constraint unit of time or T/Cu. The higher the T/Cu rate the higher the productivity. But this measure is only relevant at the system constraint.
In a manufacturing environment the constraint might be that one place whose capacity is less than the demand placed upon it. An example might include a ship loader or a train’s hauling capacity where any improvement in this constraint results in a similar (and simultaneous) increase in the systems throughput.
Regardless of its cause, the fact remains that the maximum rate of productivity for the system as a whole occurs at the system constraint and nowhere else. Therefore when management declare an intention to increase or maximize productivity (via an increase in output of the system) this can only occur by deriving a greater throughput rate at the constraint. This focus (which should really be called paranoia) recognizes that productivity lost at the system constraint is productivity lost forever for the system as a whole. Management therefore need to ensure that the system bottleneck is neither starved, blocked or subject to unnecessary downtime. This is really where the rubber hits the road because if your area of focus is not the systems bottleneck , then improvements in productivity for the global system mean something completely different. This ‘difference’ is described below.
But, what about non constraints?
Given that the vast majority (99%) of activities in a business system are non constraint activities their role in the quest for productivity improvements cannot be ignored. Therefore, if you extend the logic of maximizing productivity at the system constraint to its natural conclusion, a drive for improved outputs at a non-constraint will have absolutely zero impact on the overall systems productivity. In reality productivity improvements at non constraints often result in the unintended consequence of over investing in high levels of unnecessary work in progress (WIP) which eventually accumulate in front of the system constraint awaiting processing time at the system bottleneck. Excessive WIP absorbs cashflow, increases lead times and increases the risk of rework.
A Question of Personal Productivity
At this point it is also worthwhile touching on the subject of personal productivity initiatives
We have all heard the cries for improving our own personal productivity by such examples as reducing the frequency of email communication, sending incoming calls to voice mail, keeping meetings to a minimum and using productivity software. However, when you view these suggestions from a systems perspective do they really make sense?
Lets take an example of a business that has the normal functions of sales, distribution, production and accounting. This business has a constraint in sales. ie the market is incapable of absorbing more product despite the businesses ability to produce more.
When sales is the constraint does it make sense for people working in sales to reduce their meeting times, attend less meetings and so on? In general it does make sense if those initiatives cause those finding more sales to be distracted from that task. There will always be special cases, but lets not get too pedantic.
However now lets turn our attention to the non constraint activities in our example of production, logistics and accounting . This is where the vast majority of employees will reside and because they work in non constraint activities they by definition, have more capacity than the demand placed upon them. What is the benefit of improving personal productivity in these areas? It’s likely that people working in these areas will have more idle time as they wait for orders to generate more sales. Remember the business can’t sell more than what is ordered, accountants can’t invoice more than that which is sold, logistics can deliver more than that which is ordered. But because our scenario is to improve productivity everywhere and people are required to look busy, what does the demand for improved productivity translate into? In production it is likely to result in higher level of work in process (WIP) as they continue to produce to stock or the other functions come up with ideas to improve their productivity through faster/better computer systems or improved logistics management initiatives.
The conclusion you should be drawing at this point is that a broach based approach to productivity improvement ‘ everywhere’ is more likely to result in a negative (or at the very least, sub-optimal) outcome for the global system.
If we now turn our attention to the traditional focus of productivity improvements which can be typically described as a non discriminatory drive to reduce systems costs (inputs) we discover some equally interesting and unintended consequences.
Firstly, an across the board reduction in operating expense can result in reduced system output by either reducing the capacity of the system constraint, or decreasing the systems protective (‘sprint’) capacity at non constraints. Typical examples of where cost reductions impact the constraint would be where a cut in overtime or use of contractors for breakdowns at the constraint actually cause reduced throughput there and for the system as a whole.
In the case of non constraints, ‘sprint’ capacity is the essential part of excess system capacity needed to protect the system constraint from variation. Sprint capacity is critical when seeking to avoid constraint blockage, starvation or unnecessary downtime during high levels of variation. Without sprint capacity, the non constraints have no way of catching up with the bottleneck rate of throughput and simultaneously replenish buffer stocks required to protect the bottleneck from the next round of unforeseen events.
The sprint capacity of non bottleneck activities is a form of insurance and actually exist for these purposes – contrary to the views of the those who work in them. So whilst there will be opportunities to improve productivity via the denominator through a reduction in system inputs, care needs to be taken in ensuring that the capacity of the constraint is not compromised either directly or indirectly by compromising the sprint capacity of the non constraint activities.
I have developed a simplified set of examples of how this might work. I have used 4 scenarios to explain the point. The scenarios are 1) baseline, 2) maximize output, 3) reduce costs and 4) the system’s thinking approach. These examples are shown in the Appendix.
Based on this logic, management of non constraint activities should focus their productivity improvements initiatives in the following areas
The way forward
Continuous improvement initiatives:
Systems thinking challenges the traditional business approach to productivity by viewing a business as a system rather than a mere collection of discrete activities.
Appendix : Productivity Comparison