In a resource-constrained environment, Pareto’s 80/20 rule made all the sense in the world. When you have to make choices, tradeoffs, etc – a framework that allows you to decide what to focus on is an important tool in managing outcomes and success. But with modern technologies that simplify prioritization and decision workflows, 80/20 is a framework of the past.
For decades, an accepted truth in managing an organization has been the 80-20 rule – otherwise known as the Pareto Principle (which asserts that 80% of outputs result from 20% of inputs). The 80-20 rule made complete sense in a resource-constrained environment (particularly when data analysis was an expensive and limited resource) as it provided a framework to manage tradeoffs and prioritize focus. But with modern technologies (i.e., analytics, machine learning, and artificial intelligence) that can simplify prioritization, or can cost-effectively analyze all of the inputs, the 80-20 rule should no longer be the default in decision making.
In the past, 80-20 has been used in business to identify the inputs that are potentially the most productive and prioritize them. This would historically apply to a wide variety of business contexts:
- Which customers do I need to focus on?
- Which suppliers do I need to focus on?
- Which risks do I need to focus on?
- Which products do I need to focus on?
- Which territories, sales reps, or channel partners do I need to focus on?
- Which areas of production do I need to focus on?
- Which investments should I focus on?
80-20 was used to deal with the challenge of analyzing the amount of data to understand ‘all’ inputs (i.e., customers, suppliers, products, employees, territories, etc). It has historically been challenging, and often not cost-effective, to thoroughly analyze and understand, all of the inputs with manual analysis. If we go back a little further, before spreadsheets, most leaders did not often have the luxury of even understanding 20 percent of the inputs, which of course, gave rise to the management consulting industry. This was, of course, reinforced for me as a consultant with McKinsey, and has served me well at times throughout my career – allowing me to focus my efforts, and those of a broader organization, on what is important and should drive the desired outcome.
Today, I would challenge business leaders on their application of 80/20. With the advent of Artificial Intelligence (AI), and analytics more broadly, the cost of analysis trade-off should be re-evaluated.
Suppliers are a Good (and timely) Example
The recent supply chain upheavals in the manufacturing space have made this point clear. A small supplier, who may not be a large portion of the spend, can bring entire production lines to a stop if they have issues and there is not an easy replacement (and can even bring the line to a stop if there is a replacement if there is any lag in getting that alternative into production). Analytics and AI can be used to analyze a wider variety of factors to identify risks in a more targeted manner than just the size of the supplier.
Customers are Another Good Example
For instance, if we take focusing on 20% of customers as an example. This inevitably leads to focus on the 20% of customers that are the largest. What that does not account for is that some of those large customers are perfectly happy with you and want a low-touch relationship. Others are shifting their business and it will move away from you as a supplier no matter how much effort you put in.
Conversely, your largest customers of the future are not necessarily the same as your largest customers today – which medium customers have the potential to grow significantly? In addition, losing a large number of your “middle-60” (i.e., between 80% and 20% percentile) can also have a huge impact on your business, especially as these customers are often more profitable because they do not have the power that your largest customers have.
Today, AI can be used to sift through your data regarding your customers, their relationship with you, their loyalty, their propensity to buy from you, and what are their anticipated needs in the next period. This can allow focusing on situations with the greatest potential or on those situations that present the most risk that needs to be mitigated.
I recently had a conversation with a venture capital investment firm that was using AI to identify companies that fit their profile – certain industries, size (and using AI to estimate the size based on other data sources), growth, web traffic, etc. This enabled them to take the pattern recognition skills they had acquired through years of investing and apply that in a scalable way to available data sets and use that to focus their outreach efforts on potential investment candidates that were most likely to meet their criteria.
Some industries have already started moving in this direction. Consumer financial services use analytics & AI to evaluate transactions in real time to look for fraud and other issues. Instead of just focusing on large transactions (i.e., applying 80-20), they can analyze every transaction and focus additional effort on exceptions. Of course, eCommerce and social media have long used analytics and AI to automatically analyze each individual’s interests and then use that data to offer content or products to them that are more likely to be a good fit.
More conservative industries have been slower to leverage analytics and AI to move beyond 80-20. There are several contributing factors to this. First, having comprehensive, high-quality data as an input is not always easy, and second, the insights from that data are not always easy to discern.
For a manufacturing company, truly understanding your customers could entail knowing what equipment they have bought from you, what competitive equipment they own, what parts they have bought from you at each location, what services they use you for, how long they have warranty or service contracts, when did a channel partner serve them, which parts are relevant for their specific machines, when are they likely to need an upgrade, how long they typically go between parts/service needs, how loyal have they been, how likely are they to buy from you, what offerings are anticipated to be needed, etc. It is a complex set of data inputs and insights that are needed to move beyond 80-20.
Watch this 2-minute video to see how Entytle makes running your manufacturing company easy.
Beyond the data and insights, the other challenge is how to operationalize. Typically, more traditional industries have lacked expertise in data engineering, analytics, and AI. There are often gaps between data sets and then gaps between the data analysts and the systems that the front line uses (e.g., CRM, ERP, Field Service Management, etc). For analytics and AI to help business leaders better prioritize and make decisions, information has to be made available in an easy, accessible manner, with the output then moving towards those systems that handle the different workflows (e.g., sales, marketing, production, inventory, etc).
There are traditional companies that, together with partners, have overcome these challenges and are successfully using AI to move beyond 80-20 and focus efforts on where they can make the largest impact, not just what is most obvious. These successes have been typically driven by business leaders, supported by internal technology teams, and enabled by partners with expertise in the industry and its data model. These leaders have proven that they can take a different approach and use the data (plus analytics/AI) to find the needles in the haystack, driving productivity, efficiency, and overall better outcomes.
Also, read our article on Build vs Buy an Installed Base Intelligence Platform – Strategic Decision Framework.
Gain valuable insights through the life stories of Industry Experts, and subscribe to the Aftermarket Champions podcast.
