We live in a skewed world, where nothing is evenly distributed. For example, 80 percent of the wealth is held by 20 percent of the population. Eighty percent of your firm's sales likely come from 20 percent of its customers. And 80 percent of your breakdowns are due to 20 percent of the causes.
You may recognize this as the 80/20 principle, or Pareto's Law, named after the Italian economist Vilfredo Pareto who first recognized the pattern in 1897. Pareto was studying the distribution of wealth in England during the 1800s. Perhaps not surprisingly, he soon discovered that a minority of the population held the majority of the wealth. But when he looked further, Pareto also found the distribution of wealth was both predictably and consistently skewed, regardless of which nation or time period he analyzed.
Put simply, Pareto's Law states an almost universal truth that nothing is uniformly distributed. The comparative split between any two sets of variables may not be 80/20. It could be 95/5, 60/40 or any other variation. But it is unlikely to be 50/50, which represents linear or even distribution.
Pareto's Law has since been validated in many business applications. Pareto analysis played a significant role in the quality revolution, with two of its most notable proponents, W. Edwards Deming and Joseph Juran, applying it to identify the 20 percent of defects causing 80 percent of the quality problems.
In business sales, it is reasonable to expect that if a firm has five sales people, the top salesperson will bring in 80 percent of the sales - that's four times the sales of all the others put together.
Understanding Pareto's Law can help identify points of leverage. And the more skewed the distribution, the more powerful the leverage.
In the sales example above, the sales manager's natural inclination is usually to invest time and money to improve the results of her underperforming sales people, whereas the principle of unequal effort and return represented by Pareto's Law means a better outcome (more sales) will be achieved by shifting these resources to assisting her star performer.
Similarly, in a maintenance environment, if 95 percent of your breakdowns were due to five percent of the possible causes, it follows that if you can identify and eliminate this small percentage of causes, then you will eliminate 95 percent of your breakdowns.
As you can see, focusing available resources on identifying a small percentage of causes provides potential for exponential gains. Little hinges can swing big doors.
In the maintenance and reliability field, we already have a variety of tools designed to help us allocate resources where they are needed most. The Reliability-Centered Maintenance framework, which considers the probability and consequences of failure, is possibly the most well-known of these.
I'm not promoting 80/20 thinking as a replacement for these tools, but rather as an addition or enhancement to them. Consider this example:
During a meeting with a new client, I was briefed on the hydraulic hose maintenance program for its fleet of hydraulic mining shovels. This involved changing out every hose on the machine every 18 months. So whenever a shovel was down for planned maintenance, a portion of the hoses was changed out, beginning with the oldest first.
The hydraulic hose supplier who devised the program was somewhat of a hero because, prior to its implementation, improvised hose replacement in response to in-service failures had resulted in machine availability falling to as low as 65 percent.
When a multimillion-dollar shovel stops, so does a multimillion-dollar fleet of haul trucks. So downtime is a major cost. But large-diameter, multispiral, hydraulic hoses aren't cheap either.
I couldn't argue with the success of the hose replacement program; however, I did point out its fundamental flaw. It treated all hoses equally. The 80/20 principle suggests it would be highly unlikely that 50 percent of the hoses were responsible for 50 percent of the in-service failures and downtime.
So I explained to my client if he were to look at the historical data, he should expect to see a skewed picture: that 20 percent of the hoses were causing 80 percent of the in-service failures and downtime.
In fact, the available data revealed less than 20 percent of the hoses were responsible for nearly 90 percent of the failures (note that we are comparing two sets of unique data, so the comparative split doesn't have to add up to 100).
This discovery not only reduced my client's hose bill (oils spills, ingression of air and other contaminants), but the risk associated with the introduction of human agents as a result of unnecessary hose replacement was eliminated as well.
As this story illustrates, if you are applying linear thinking to any facet of your maintenance strategy, it's wrong. An 80 percent reduction in your maintenance costs will likely come from 20 percent of your maintenance effort. The trick is in identifying the significant 20 percent.