By: Brenna Hoffman
Most people in the world have experienced the incessant questions that a toddler asks: Why is the sky blue? Why do I have to go to bed? Why is that man naked? From the time they wake up to the time they go to bed, toddlers are obsessed with asking “why” for almost everything they see. The questioning annoys parents and guardians alike, but the questioning skills of a toddler are actually very important in the professional world.
Managers need to ask questions. More importantly, managers need to ask the right questions to find the root cause of a problem[1]. Unfortunately, most managers do not understand what to ask or how to find the right questions. One way to find the “right question” is to look for the root cause of the problem. However, one the root case is identified, the battle is not over. The root cause of the problem has to be fixed.
This blog will explain one way to ask the right questions to get to the root cause, the 5 Whys, then solve it by using Systems Thinking. These approaches are important for data scientists as well as managers. In order not to waste time or resources, it is important to diagnose the root cause of the problem; the 5 Whys can help data scientists work with their business users to refine their questions to get to the root cause. Then, Systems Thinking will empower them to effectively solve it.
Identifying the Root Cause of the Problem: The 5 Whys
The goal of the 5 Whys is to get to the root cause of the problem by starting with the problem statement (i.e. My dog is hungry) and then asking “why” several times to surface the actual root cause of that problem (i.e. I was too lazy to go to the store to get food). Identifying the root cause is critical: rather than addressing just the symptoms of the problem, the root cause is something a manager can take action on to fix or stop the problem. Asking why five times is usually sufficient to get to the root cause.
According to the website, ISIXSIGMA, this method is a valuable approach when “problems involve human factors or interactions.”[2] Eric Ries also “posits that behind every supposedly technical problem is actually a human problem.” Hence, using “The Five Whys for Start-Ups” can be valuable.
In addition to providing value for problems that are based on human interactions and issues, the 5 Whys is also easy to understand and does not require statistical tools.
However, this method can have a few potential pitfalls. First, according to Adam Ramshaw’s article “5 Whys Root Cause Analysis Template and Process,” make sure to not get into a loop. Some questions may lead back to the original problem statement instead of the root cause. Second, some people may mistakenly believe they have reached the root cause when they should still be asking questions. These pitfalls can be avoided by making sure the final answer is an actionable item and to move one question back if one gets into a loop.
The 5 Whys is the used to identify the root cause of a problem. Once the 5 Whys have been used to surface the actual root cause of a problem (i.e. know what the right question is), then the next step, systems thinking, can be used to help solve it.
Solving the Problem by Using Systems Thinking
Systems thinking is “an understanding of a system by examining the linkages and interactions between the elements that comprise the whole of the system.”[3] Donella H. Meadows identifies different leverage points in a system to change the system. . Leverage points “are places within a complex system (a corporation, an economy, a living body, a city, an ecosystem) where a small shift in one thing can produce big changes in everything.”[4] “Places to Intervene in a System,” published in Whole Earth Winter in 1997, explains in detail the nine leverage points, shown in Table 1.
For example, most drivers have probably used a crowbar. a common tool that uses leverage to break or separate things. The longer the crowbar, the more strength it has and the more leverage that can be applied. Meadow’s leverage points are similar to the crowbar in that the higher they are on the list, the more leverage and strength they have to change the system—but also but also the more resistance to change they will encounter. For example, the first two leverage points (numbers and material stock and flows) are the physical parts of the system. They are harder to change physically but affect the system less.
In addition, Meadows warns that many companies push leverage points in the wrong direction. This is because complex systems are hard to understand and may work in a counterintuitive fashion.
This blog will not touch on all of the leverage points. Instead, I will touch on the following points that resonated most with me: regulating negative feedback loops, the rules of the system, the power of self-organization, and the mindset or paradigm out of which the goals, rules, feedback structure arise.
Table 1: Meadow’s Nine Leverage Points[5]
| Leverage Point | Definition |
| Numbers (subsidies, taxes, standards) | These are the parameters of the system. |
| Material stocks and flows | This is the layout of the system. |
| Regulating negative feedback loops | A loop that self-corrects if the situation varies from the goal. |
| Driving positive feedback loops | A loop that is “self-reinforcing” and will continue to grow in power. Meadows used the following example: the more babies are born, the more people grow up to have babies, the more babies are born. |
| Information flows | This is where the information is shown or who the information is shown to in a system. |
| The rules of the system (incentives, punishment, constraints) | The laws and controls that govern the system. |
| The power of self-organization | The ability to change the system under certain conditions. |
| The goals of the system | The desired results of the system. |
| The mindset of paradigm out of which goals, rules, feedback structure arise | The shared ideas of the system. |
According to Meadows, a negative feedback loop is a function to “keep system states within safe bounds.”4 These loops need “a goal, a monitoring and signaling device to detect excursions from the goal, and a response mechanism.”4 The more unambiguous a system is, the more corrective and powerful the loop will be. Meadows provides an example of a thermostat keeping the temperature of a room stable. The thermostat has a goal (keep room at a certain temperature like 70 degrees), a monitor/signal device (a sensor to say what the actual temperature is), and a response mechanism (if the temperature of the room is lower than 70 degrees, the heater will turn on; if it is higher, the air conditioner will turn on). However, if all of the windows are open, then the thermostat’s corrective power decreases because ambiguity increases.
The rules of the system include incentives, punishment, and constraints. The rules “define [the system’s] scope, boundaries, and degrees of freedom.”4 Meadows states that “if you want to understand the deepest malfunction of systems, pay attention to the rules, and to who has power over them.” The right rules can change behavior if they are enforced. I believe that if leaders and other influencers follow the rules then the entire way a system works can be changed drastically.
Meadows states that “self-organization means changing any aspect of a system lower on this list.” This could include changing physical structure, positive or negative loops, etc. She states that it is important to write “rules for self-organization…[that] govern how, where, and what the system can add onto or subtract from itself under what conditions.” In other words, they are rules written by the users of the system at the beginning that explain what can be changed, when it can be changed, and how it can be changed. Self-organization can be helpful because it allows for an adaptable system under certain guidelines. Instead of having an immutable system, which could prevent change when needed, self-organization gives the “ability to survive change by changing.”4
The last point to be discussed is the mindset or paradigm out of which goals, rules, and feedback structure arise. A mindset or paradigm is the “shared idea in the minds of society… Everyone knows them… [they] constitute the society’s deepest set of beliefs about how the world works.”4 They are the sources of the system. Mindsets can come from the leaders of the company, but the goal is to have them be embraced from everyone. Even though they are the easiest and cheapest to change physically, they are the most resistant to change from individuals because it involves changing the way they think.4
Meadows offers some words of wisdom to people who are trying to use these leverage points to solve problems. First, this is “an invitation to think more broadly about system change.”4 Second, her leverage points are not a step-by-step list. Managers should not go through the list one by one. Instead, she suggests trying to figure out which leverage point will help change the system and can actually be applied given the circumstances. Third, try not to get attached to paradigms. They change, and not a single one of them is the only source of truth. Finally, Meadows states that “leverage has less to do with pushing levers than it does with discipline thinking combined with strategically, profoundly, and madly letting go.” This last part is key because strategic thinking is important in any organization.
Now that both of these processes have been defined, next I am going to explain how they can work together to solve a problem.
Using the Two Methods
While the 5 Whys focuses on getting to the root cause by asking the right questions, Meadow’s systems thinking approach focuses on different places (called leverage points) to intervene in a system to change it. These two methods can be used in conjunction with each other. First, identify the problem statement; use the 5 Whys approach to get to the actual root cause of the problem, not just the symptoms. Once the root cause has been agreed on by the people affected, use the systems thinking approach to find different potential leverage points in the system to help solve this root cause.
To help solidify this process, I applied this approach to a sales company that I know. A sales company, called BIG[6], has been paying its salesforce incorrectly. The payments generate automatically from a commission system that has been around for about five years and the salesforce are on a tiered commission plan.
Figure 1 shows the 5 Whys approach to finding the root cause of paying people incorrectly. The problem statement is “Payments are Incorrect.” After a series of questions, the root cause was finally revealed- the programmer does not understand the business model, and the user does not understand the complexities of coding. While I was creating this 5 Whys diagram, I originally stopped at “Because there is a communication barrier.” However, after remembering that this was one of the cautions, I asked one more question and found the true root cause that can be worked on to solve the problem.
Using the leverage points from Meadow’s article, there is one main leverage point that can be used to solve the root cause. There are also other leverage points that can be used to keep the problem from happening again.
The main leverage point is the power of self-organization. Stated above, self-organization is the ability to change the system under certain guidelines. BIG allows for changes made to the system. To this regard, the commissions system would have to be changed. The first step is for the business user and programmer to sit in a room together until both understand the complexities of each other’s world.
The next step would be to delete incorrect parts of the structure (aka the code), and re-add new parts, making sure to test the entire time. This could take some time but if it prevents incorrect payments, then it would be worth it. After the system has been recreated, two other leverage points could be used to make sure payments stay correct: negative feedback loop and rules.
A negative feedback loop could be added in two instances. The tiered commission plan is supposed to be recalculated nightly. A negative feedback loop could send an email stating it has not been run correctly. The email could also have a button that will allow the user to run it manually. The second instance could be another loop where a list of test cases can be constantly monitored when code is changed. It could alert the business user when a test case is not paying like expected. The second leverage point that can be used is to create a new rule that either punishes or incentivizes the commissions team, including the programmer, when a payment is either correct or incorrect. BIG would have to decide which way would work best for the commissions team.
Conclusion
The 5 Whys and Systems Thinking combined approach can be beneficial to a company if used correctly. Before reading about these two methods, I always thought the problem statement was the root cause. I did not realize the problem statement is usually just a symptom of the root cause. I talked to the programmer at BIG after finding the root cause. We have also talked about how to solve this problem going forward instead of just adding new code to fix the symptoms every time. We talked through the main leverage point I mentioned above and have decided to rebuild most of the commission system. Since we are just starting on rebuilding, I cannot say for certain if this leverage point will work, but it is worth a chance to fix the root cause.
Understanding the actual reason a data scientist is doing something for a company (e.g., the problem they are being asked to solve) is important to their success. The method discussed in this paper can help data scientists understand the why behind requests by helping users refine their questions to get to the root cause. Once, the root cause is known, the data scientist can use Meadows’ systems thinking (or a different approach that works better for them) to solve the problem. In the future, I am going to use this method as a data scientist to make sure that I have always found the root cause and to identify leverage points in the system that can be used to fix the cause.
References:
“Determine The Root Cause: 5 Whys.” ISixSigma, 27 Nov. 2018, www.isixsigma.com/tools-templates/cause-effect/determine-root-cause-5-whys/.
Meadows, Donella H. “Places to Intervene in a System.” Whole Earth, 1997.
Repenning, Nelson P., et al. “The Most Underrated Skill in Management.” MIT Sloan Management Review, MIT Sloan, 13 Mar. 2017, sloanreview.mit.edu/article/the-most-underrated-skill-in-management/.
Ries, Eric. “The Five Whys for Start-Ups.” Harvard Business Review, Harvard Business Review, 23 July 2014, hbr.org/2010/04/the-five-whys-for-startups.
“Systems Thinking.” Learning for Sustainability, learningforsustainability.net/systems-thinking/.
[1] Repenning, Nelson P., et al. “The Most Underrated Skill in Management.” MIT Sloan Management Review, MIT Sloan, 13 Mar. 2017
[2] “Determine The Root Cause: 5 Whys.” ISixSigma, 27 Nov. 2018
[3] “Systems Thinking.” Learning for Sustainability
[4] Meadows, Donella H. “Places to Intervene in a System.” Whole Earth, 1997
[5] Donella has changed her nine leverage points into 12 leverage points. The original nine are discussed in this article. The link to the new points is http://donellameadows.org/archives/leverage-points-places-to-intervene-in-a-system/
[6] Name has been changed.
MSBA 