Systems diagnosis tools and techniques – Part One

Systems diagnoses and solutions are both the answer to a lot of what is wrong with corporate performance – and a source of the problem at the same time. Most issues around strategy execution and organizational performance have foundations that spread across different parts of the system. Yet way too often, the solutions that are tried are too narrow and fail to properly address the true root causes and interdependencies that are endemic. Too many approaches to improving performance fail to take enough of a systems view, or apply it in the right way.

When I wrote my book Strategic Analytics eight years ago, this was the fundamental problem I was focused on: how to get people to ask the right questions and apply the right kind of diagnostics to identify the right solutions. I was hardly the first, and certainly won’t be the last, to call out deficiencies with how we approach finding the right solutions to organizational performance.

The history of such tools is long and storied, and includes:

• Deming’s focus on quality and systems tools, which became the foundation for total quality management and the Toyota Way
• Design thinking
• Agile
• Six Sigma (and Lean Six Sigma)
• Business process reengineering

Comparing and contrasting these different approaches requires an entirely different article, for another time. Here I’m going to address the common threads in these approaches, which span over seven decades of development and implementation.

Systems diagnosis is one of the most important, yet greatly underutilized, tools for improving organizational performance and strategy execution. It is both everywhere – as evidenced by so many tools that have been developed over the years – and glaringly lacking at the same time. Lacking because, despite so many people having been trained in various methods over the years, day in and day out, people in key leadership and project management roles fail to be as comprehensive as they need to be in carrying out their work.

Take a broad and comprehensive view of the issue

Beware the leader who is convinced they know exactly what the solution is to the problem. Watch out for competing objectives and conflicts among leaders in different parts of the system. Zoom in and out. Look left and right, up and down. Be aware of interdependencies among the parts of the system. Focus on perennial problems that never seemed to get solved fully in the past. Check for unintended consequences of what you want to change.

These are some of the coaching guidelines that are common among systems diagnostics. If you do all of them, you will greatly increase the chances you will get to the bottom of what’s challenging your organization’s performance.

In this first installment, I will address the first half of these guidelines.

Beware the leader who is convinced they know exactly what the solution is to the problem. Watch out for competing objectives and conflicts among leaders in different parts of the system. These two points are related. When a leader comes to you and is absolutely sure they know not just the problem to be addressed, but which solution to apply, more often than not, they are at least a little incorrect. They usually have failed to grasp all the potential root causes for the performance problem. So while their identified solution may be appropriate to apply, additional work is almost always needed to make sure that is the correct solution, and whether other changes are needed at the same time to fully address the issue.

Part of the reason why you cannot just trust what the leader tells you is that, unless they are the CEO, they do not have responsibility for the complete end-to-end business processes that cross all departments, functions and business units. Because they are responsible for only a subset, chances are they are looking at the issue with too-narrow a focus that takes into account what they personally are responsible for, and downplays what their peers in other parts of the business are responsible for.

Zoom in and out; look left and right, up and down. The focus here is on the same thing: making sure you are not looking too narrowly or widely when diagnosing the performance problems.

This guidance can be traced back to the work of systems diagnosis pioneer W. Edwards Deming who was trying to solve the problem of how to improve quality in manufacturing. When he was doing his early work in the 1950s, a main challenge was reducing errors in manufacturing assembly lines – the ones that had been created a half century earlier by Henry Ford and Frederick Taylor.

The initial assembly lines created by Ford and Taylor were modern engineering marvels, and they set the industry standard for decades and helped propel the industrial revolution. Yet over time it became apparent that consistent quality was difficult to achieve under that work design. To determine the sources of the quality problems, Deming had to “zoom in” to look at how the work was happening at the job level across all the different jobs on the assembly line. After zooming in to the job level, he then “zoomed out” and asked the question, if the work were reorganized from narrow jobs into self-managing teams responsible for a larger part of the work, would that solve the quality problems? (The answer, as history has demonstrated, was a resounding “yes.”)

Note that zooming in from the product level to the job level is like “looking down” from the product to its subparts. Zooming out from individual jobs to multiple roles that assembled a compete subpart of an automobile, such as the engine, is like “looking up” from individual jobs to groups of jobs. So “zooming in” and “looking down” are the same, just as “zooming out” and “looking up” are.

“Looking left and right” is a way of pointing out that any one job or any one team does not operate in a vacuum. The job/team has to integrate with others: the jobs/teams that do the work immediately before and after. So if you are examining issues that appear to be rooted in one job’s performance, you need to understand how the work of that one job integrates well (or not) with the work that came immediately before and which follows immediately after.

Be aware of interdependencies among the parts of the system. Looking left and right at the workflow addresses interdependencies of the work with adjacent parts of the system that come immediately before and after. Beyond those immediate interdependencies, there are others if you zoom further out to look at the entire end-to-end business process.

Pushing on a balloon. Playing whack-a-mole. These are two metaphors that call out potential interdependencies in the system.

When you push on a balloon, whatever part you are not holding directly will bulge or expand out – you cannot push on all sides of the balloon simultaneously to prevent the bulging. And you can never really accurately forecast which parts will expand. Whack-a-mole has a similar connotation: you never can tell where the next problem will arise.

The power of systems thinking or diagnostics is that you take a step back to consider why things are happening the way they are. And usually there are interdependencies that can help explain which parts of the balloon will bulge out – and where on the playing field the next mole will pop up in the game of whack-a-mole.

For example, the recent supply chain travails created by the Covid-19 pandemic are a classic case requiring systems diagnostics to identify where the bottlenecks and shortages were created, and identify options for overcoming them. One of the first things that many consumer products companies pivoted towards was producing a much smaller number of options for products which, over the years, had grown in scope and complexity. So rather than have, say, 17 different shades of white paint (or variations of shampoo or salty snacks or chocolate), a company might pare the options back to the 3-5 which traditionally were the biggest sellers.

The benefits of paring back product options were many:

• Fewer products options produced, which simplified operations, reducing the number of issues that had to be dealt with internally, in an environment that suddenly had become much more uncertain. This made the jobs of staff simpler, which was important in an environment where employee absence and turnover risks had suddenly increased dramatically.
• The simplified operations included dealing with far fewer external vendors and suppliers of the inputs needed to produce their products, which lowered the complexity of dealing with disruptions that were happening at potentially every point in the supply chain. This included reduced cost of carrying excess inventory of both inputs and finished products, compared to how much extra inventory costs would have been needed to maintain the pre-Covid full lineup of product options.

The system-level interdependencies here included:

• Disruptions in the operations of vendors and suppliers due to their labor shortages
• Disruptions in the distribution system for the inputs from vendors and suppliers, and for the outputs of the company’s own operations
• Internal labor and product disruptions at any point within the internal operations of the company, which would ripple throughout the entire system

The rest of the systems diagnosis coaching guidelines are addressed in Part Two.

This content and principles of this article are a key component of the virtual People Analytics and Change Masterclass, which starts on January 31 and runs through February 16. The workshop is ideally suited for HR Business Partners and others who want to improve their ability to address the issues that really matter for improving organizational performance ... and ensuring that the changes made are the right ones and will have lasting impact.

#systems #systemsthinking #leadership #HR #organisationaldevelopment #SWP #HRBP #OD #changemanagement