The pitfall of 80/20: Think about it!
Think about it!
The 80/20 rule (also known as the Pareto Principle), first founded by the Italian economist, Vilfredo Pareto in the area of socioeconomics. It was adopted by Lean and Six Sigma gurus then widely engraved in most professional practices in business and beyond. By principle, the 80/20 rule drives us to pay attention to the little portion of anything which drives the big portion of outcome. This was illustrated in Pareto’s example: 80% of the wealth is owned by 20% of the population.
80/20 beyond socioeconomics
The 80/20 rule is fundamental to everything we do in our personal lives, business and beyond. Sales people might prioritize customers based on the deal size; marketers prioritize customer segments based on the lifetime value; accounting and finance professionals priorities accounts based on their weight of the overall; and Human Resources professionals prioritize departments based on their headcounts (or maybe not).
Whether these scenarios apply to all or not (definitely not), the point is that we’re often misled by one metric and we overlook the other ones which are also vital to the outcome we are trying to control. In other words, we tend to focus on the first layer of a quantifiable metric and take things as they are. While we tend to focus on significance, in many cases, we overlook the fact that there are different dimensions through which we can evaluate significance. In this copy, I provide exaggerated examples to make clear how this is relevant in different business contexts.
Potential pitfalls and counter views
In business and life in general, we cannot solve all problems, but we can prioritize them. In manufacturing quality control for example, focusing on solving the issues which cause the most product defects is one way this concept is applied. In other manufacturing areas, process improvement focuses on leaning out the process from barriers (disrupters) of product flow. But that’s too simplified; because how can we judge which barriers are the most problematic? Some sort them by frequency of occurrence, others focus on the overall time lost or maybe cost of material? Well these are all different ways to look at it and are definitely great, if aligned with business objectives and priorities. We must always have the end-result and objective in mind.
Speaking of this, imagine you were a manufacturing or production manager and wish to improve operations and productivity. So, you track the downtime and categorize it based on the type of problem causing it (a common thing to do). By end of month, consolidate the data and you’ll have a table like the one below showing the problem categories and the relative downtime in hours for each throughout month:
The typical approach would be to prioritize machine downtime as something which needs to be resolved first/immediately, while other issues can wait, for they aren’t as significant. Now let’s say we conduct a workshop and focus only on the machine downtime, for it has attributed to a total of 8 hours during the month, the highest amongst problem categories, with an assumption that one solution could resolve all machine related issues. Because after all, we’d want to do the least to resolve the most; the basic 80/20 principle would guide us in that direction.
However, what if for once, we lightly explore the other problem categories, and this time with the potential solution in mind, even at a high level. Because eventually, we could end up trying to resolve many different sub-machine problems, which are causing machine downtime; it could be that the causes of machine downtime are fragmented across a different set of root-causes; which means, we could end up solving many problems or launching several projects to do so. So, exploring and scanning the other downtime problem categories (tools, material, etc.) first, could help us identify potential solutions and see crosscutting opportunities in which one solution type (e.g. conducting training) or project could address multiple problems (or problem categories) at the same. If we pause and think about it: Isn’t this 80/20 too? It’s just how we choose to see it.
In this scenario, the typical categorization dimension segments downtime based on the type/category of problem; another segmentation in terms of possible solutions could prove extremely useful, especially in cases where the significance of problem categories is similar and where problems are not too complex to resolve, and understanding them doesn’t require a lot of time and resources.
To further illustrate, the table below includes additional columns, which represent the downtime broken-down further, by type of potential solution for each of the respective problem categories. In this case, imagine when analyzing the causes of machine downtime, we realize that 5 out of 8 hours of machine downtime is caused by issues which could be addressed/resolved by some training, likewise some of the downtime that is attributed to tools and materials. This leaves us with a total of 9 hours of downtime to address by implementing a training program; wouldn’t this be a good idea? I know, we don’t have the full picture, but you know…
This breakdown helps us see what’s significant from another dimension, with the solution in mind. In reality, just investigating problems alone could take a lot of resources, however, in many cases, it doesn’t. It’s not just about problem vs solution, but really the emphasis on prioritizing (80/20) based on breakdowns or dimensions which are most linked to the actual efforts and outcomes, starting with the end in mind. In a strategy or business plan, clustering of initiatives could be done a certain way for comprehension and coherence with other parts, but for decision-making on phasing and implementation, different ways of clustering could prove more useful. Should you launch one program in all offices, or all programs in one office?
More examples in other domains
Think about it: If you’re developing your growth strategy and considering investing large amounts of cash to grow your business. ‘Where you invest it’ is one question, ‘how you invest it’ will come right after. This brings up another question around how you should prioritize or order your investments, and many factors come into play. You could prioritize your projects from an ROI perspective or you could do so from an implementation/execution perspective or other synergies. I mean it’s not always just about which projects bring the highest ROI, but also which projects can share the same cost or executional efforts altogether, and of course there would be other dimensions which would drive how we should cluster and group projects. This point here is that holistic thinking in this domain would incorporate more than one dimension.
Now let’s say we’re looking at the income performance of a corporation which has many business units. A common approach, and I’ve seen it, would be to look at revenues or profits of the different business units and maybe the margins, sort them and focus on the best-looking ones to promote and worst-looking ones to improve. Would you argue that this isn’t the best approach? I would.
I’m not against looking at the metrics this way, but I have an issue with ONLY doing this. While this view shows us the current performance of a business unit, we should be able to know much more before making business decisions. For example, the absolute and percent change over time will tell us a lot about the trend (where there’s abnormal performance); the change and percent change of a business unit’s contribution to the company’s top/bottom-line, the cost contribution of each business unit to the whole corporation, the absolute and percent change in cost contribution of each business unit, the absolute and percent change in costs and more, each will tell us something, and it’s not always the absolute value of the current performance figure.
These aren’t new ways of analyzing performance, neither are they comprehensive, yet although they provide valuable insights, they’re often ignored. This is another example where the 80/20 rule could fail us, or we could fail it, by limiting its use to just the obvious. If we were only evaluating revenues, should 80/20 be applied by evaluating the performance of different products/departments, or should we also segment customers and look for potential trends among other metrics? We’d need to think multi-dimensionally.
If you’re evaluating customer service and 80% of your complaints come from 20% of your customers, would you focus on resolving the complaints of only these customers? What if these customers are just costly to your business and have all kinds of complaints? Later, you learn 90% of all the complaints are attributed to one issue that is common across all customers? Would you resolve all the different issues coming from 20 out of 100 customers, or resolve that one issue which is repeatedly causing 90% of the complaints? I’d go for the latter. However, if I’m also given some information about the Life Time Value of the different customers, I’d probably think again. We shouldn’t be one-dimensional.
Efficiency vs effectiveness
All in all, we tend to split/segment things one way and prioritize the biggest segment, assuming this would help us effectively solve the problem at hand, and considering we’d be more efficient in solving it. However, we can make something efficient, but we’re only effective if we know what is best to make efficient. Just like choosing which processes in the supply-chain to fix first. Effectiveness always goes one step further; and rather than being only attributed to output (like efficiency), it is more about the outcome.
While efficiency is about doing things right, effectiveness is about doing the right things. Why would doing something right matter if it’s not the right thing to do after all?
Way to go…
To conclude, I meant to argue that while it can be considered effective to follow the 80/20 approach, it can be misleading if the dimension/metric being considered is not sufficient. We could only be effective if we pay real attention to which dimensions/metrics really matter, beyond the obvious. Sometimes the important metric is intangible, or cannot be quantified but is of critical value or relevance to the business outcomes.
If you’ve found this to be useful and/or would like to stay up-to-date with upcoming thoughts, you can subscribe to receive notifications and future updates.
