Proving Productivity - Measureology

Productivity. What is it? Can it be measured? Should we bother?

Language Barriers

Language is a big obstacle when trying to measure things.

Some organisations love important-sounding words: Efficiency, effectiveness, sustainability, transformation, delivery and so on. The word ‘productivity’ is no exception. In some fields such as manufacturing or a nation’s economy, productivity may be well understood. In other settings, the concept of productivity is vague, woolly, slippery.

To measure anything we must be able to observe and detect it in the real world. We can only do that if we can first describe that thing without ambiguity. Whether or not we can then collect data for that measurement is a different challenge.

Rushing to measuring stuff can bypass a common understanding of its decision purpose. Using available data sources may seem like a short cut but may yield less useful information. Neglect this, and trivial metrics could drive the wrong behaviours and provide weak feedback of change.

Maybe if we get to know productivity a little better we can understand why, and how we might measure it.

Labour Productivity

Perhaps the most pervasive uses of the word ‘productivity’ in our daily lives here in the UK are the quarterly reports of economic performance from the Office of National Statistics. These might appear as an attention-grabbing headline like : ‘UK productivity suffers worst drop in five years’.

This form of productivity is termed Labour Productivity which is a ratio of an economic measure of output called Gross Value Added (GVA) divided by a labour input in terms of the number of worker hours.

Higher economic output, whether GVA or GDP, is always communicated as being ‘good news’. Increased productivity, with GVA in the numerator, inherits this apparently virtuous goal too. Productivity growth is good; bigger is better.

Productive Production

The origins of Labour Productivity are grounded in the creation (production) of physical goods (products).

For most of human history, producing food or goods to sell for food were tightly bound to the number of hours of labour needed. Labour-saving tools and then horses were harnessed to reduce labour hours and their physical effort.

In the 1700s, steam power began to replace horsepower in taking the strain. Better tools, the machines of the industrial revolution, decoupled production from the direct time and effort of human labour.

Machines made it possible to create the same output with fewer, less skilled and lower paid workers. Company owners and their investors reaped the rewards of this technological revolution. The workers, maybe not so much.

Expectations of perpetual economic growth through increased productivity had begun. Use of the term ‘productivity’ in books hit its stride in the early 1980s and remained present for another 20 years.

Even now, modern technological advances such as robotic automation are claimed drivers of increased productivity through reduced labour hours. Much has been written too about the ‘Productivity Paradox’, the absence of economic growth in spite of advances in Information Technology.

Personal Productivity

Another common reference point is our own personal productivity.

Back in the 1990’s we called it Time Management but the personal improvement industry has moved on since then. Now, gurus like David Allen, Charles Duhigg and Graham Allcott claim they have the secret hacks which can turn us all into Productivity Ninjas.

This form of productivity aims to make the best use of every minute of our day to make progress towards our most important work and life goals.

Fundamental Equation

The dictionary definitions of productivity can help too.

There are simple definitions such as: ‘The rate at which goods are produced’. There are the more elaborate like ‘The rate at which a worker, a company or a country produces goods, and the amount produced, compared with how much time, work and money is needed to produce them’

At its broadest, productivity relates outputs to the inputs required to create them. Productivity detects how much of an input to a system process is consumed to create an output, a form of conversion or yield.

The general form is:

Productivity = [Outputs]/[Inputs]

As a rate, the numerator and denominator are often in different units. Common forms of productivity describe the rate of output per unit of time, whether worked or elapsed. GVA per Worker Hour, Tasks Completed per Day.

Its common to think of productivity as the throughput of work; the volume of jobs or items completed per unit time. Here the focus is on increasing the amount of work flowing, with perhaps less focus on the composition of that work.

Productive or Efficient?

But wait! But how is productivity different to efficiency? According to Wikipedia, ‘Efficiency is a measurable concept, quantitatively determined by the ratio of useful output to total input.’

Efficiency = [Useful Outputs]/[Total Inputs]

The important word here is ‘ratio’. Efficiency, as a ratio, is usually expected to have the same units in the numerator and denominator.

Efficiency might be distinguished as seeking the highest amount of output for the least amount of input. The higher output numerator could be a ‘higher volume’ or ‘higher specification’. The lower input denominator could be money or a form of physical resource but could be time too. Efforts to increase efficiency might focus on creating a ‘better’ output, on eliminating input wastes, or both.

So productivity and efficiency may mean similar things, or they might not. The semantics can get a bit tangled, with different uses in different fields.

Both productivity and efficiency express the desire for ‘doing more with less’. Some draw a distinction between productivity as ‘doing more things’ and efficiency as ‘doing things better’. In almost any continuous improvement setting we’re hoping to do both, and the specifics of the process at hand matter more than the semantics.

If anything, this close coupling of productivity and efficiency further underlines the need to use very specific language to avoid confusion. The test of a good measurement might even be whether we’ve eliminated the words ‘productivity’ or ‘efficiency’ altogether.

Why Measure Productivity?

Just because we can measure productivity doesn’t mean we should. Again, the basic equation gives us a clue to its potential value.

Inputs are scarce resources which can be boiled down to a commitment of time or money. It makes intuitive sense to ask how much output of completed work we are getting in return for that commitment. Productivity would seem to have a very obvious improvement goal: we want more work done for our committed input.

To suggest that productivity growth isn’t intrinsically good seems heretical. But as all Systems Thinkers know, there may be limits to growth somewhere in any system. Or that a preoccupation with productivity in one part of a system is likely to harm another part of it, or an external system.

It might be that the best we can hope for is improvement towards a limit as well as having a balancing measure.

Productivity Example

Here’s a great example from Stacey Barr which illustrates using two balanced measures of productivity for marketing content creation.

The first measure relates the number of individual items of content created to how long it took to create them. This is the raw volume of product at the output of the content creation process.

Content Creation Productivity = [Content Created] / [Content Creation Time]

Content Created: the number of unique content packages, such as articles or videos or podcasts, produced each month.

Content Creation Time: the number of hours spent creating content. The hours spent publishing & promoting could be added to this.

Content Reach Productivity = [Content Reach] / [Content Creation Time]

Content Reach: the number of times any item of content that is published is consumed (e.g. downloaded, watched, opened, etc…)

The numerator here isn’t just a count of the content but the consumption of that content. This is a powerful measure because it looks beyond the ‘volume’ of content to the ‘value’ of that content to a consumer and relates that back to the time taken to create it.

This example is instructive.

It is internally focused on a particular Content Creation process with an improvement intent.
There are two measures in tension, so we don’t just drive volume without thinking about customer value.
This might lead to a decision to shift the type of content creation; reducing volume but improving reach.
Reducing the productivity of the first measure might reduce waste and increase efficiency in the second.
A process is being measured, not a person. But people in the process can understand how they are personally contributing.
There’s likely to be local influence over how to improve the Content Creation process and use the measures for feedback.
This information has zero value for the manufacturing team, who have their own widget productivity measures.
It may not be make sense to compare, benchmark this measure with any other content creation units but we do want to see how this changes over time.
We could dispense with the word ‘productivity’ altogether and refer to a ‘Content Packages Created per Hour’ or a ‘Content Reach per Creation Hour’.
The Content Creation Time could be replaced with a Content Creation Cost to communicate a rate of return.
Without this measurement, we wouldn’t know whether were getting more bang for our Content Creation process buck.

PRODUCTIVITY Principles

Here are a few productivity measurement principles worth considering.

Productivity can have internal or external viewpoints. The internal viewpoint detects whether a system is getting better or getting worse. The external viewpoint compares multiple systems.
The value of measuring ether form of productivity will depend on which decisions are made better with that information.
For the interval viewpoint, productivity improvement goals will be different for each system. We should expect the measurement to be different too.
Each productivity measurement may have its own purpose depending on the decision audiences and intent. Useful information for one audience may have zero value for another.
Focus productivity measurement effort first on those parts of the system you want to explicitly improve, or those parts where erosion of productivity would cause system decline.
Measure the productivity of processes, not people. Measuring the productivity of people is a sure way to encourage creative ‘gaming’ by people constrained by a process.
There may be many forms of productivity in each system because inputs and outputs look different in each process.
The inputs and outputs will ideally belong to the same process or causal chain of processes so there’s some scope of influence over productivity improvement.
Looking downstream at the causal results of a process output can capture a richer picture of productivity than just counting the intermediate outputs.
There will often be a tension between different measures of productivity. Focusing on just one may produce a systemic distortion when what we really want is an optimum balance.