Logic tree root cause analysis: A method to have productive disagreements.
Summary: If you disagree with someone, it is an opportunity to learn something! One way to figure out what there is to learn is to do ‘logic tree root cause analysis’.
Logic tree root cause analysis steps:
1. Decide on the problem statement (hypothesis).
2. Build and agree on a Logic Tree by MECEing the problem statement.
3. Find out where in the Logic Tree you disagree (eg you view good, they view bad). IE find the root cause of disagreement.
4. Do the work to find out which viewpoint best reflects reality at this point.
Example: How can we increase profitability?
Person A thinks we should build a new feature.
Person B thinks we should decrease customer churn.
What can you do? Logic tree root cause analysis :)
1. Decide on the problem statement (hypothesis).
How can we increase profitability?
2. Build and agree on a Logic Tree by MECEing the problem statement.
This part can be quite hard… but hard can be fun!
3. Find out where in the Logic Tree you disagree (eg you view good, they view bad). IE find the root cause of disagreement.
Person A thinks we should build a new feature.
Person B thinks we should decrease customer churn.
You end up going one layer deeper to have ‘better onboarding’ and ‘usage based pricing’.
4. Do the work to find out which viewpoint best reflects reality at this point.
Person A thinks that a new feature will double conversion from trial to paid but Person B thinks it won’t make much of a difference.
Both Person A and Person B think that better onboarding will decrease churn by 20%.
Outcome:
It is decided that a quick low resolution prototype of the possible new feature will be built and discussed with 5-10 trialing customers for their feedback and then the results reported back so a high quality decision can be made.
This should mean that you can have much better confidence around the best way to ‘improve profitability’ rather than just either going with ‘new feature’ or ‘reduce churn with better onboarding’.
Recommended not to do.
The most stubborn person prevails.
Go with the HIPPO (highest paid person's opinion).
Build a defective Logic Tree and as such don’t identify the root cause, but a proximate cause.
You don’t learn from people who agree with you.
You don’t have battles of wills, with the most stubborn prevailing.
“The good learn from everything and everyone, the average from themselves and the stupid already know everything.” - Socrates.
Find the root cause of where you disagree and then do the work to figure out what best reflects reality at the root cause.
Jingle: I don’t want to be ‘right’ the most, I want to be the one who learns the most.
Related blogs:
+++++++++++++++
Details
Proximate vs Root Causes: Keep Digging to Find the Answer
Proximate Cause
Definition: The proximate cause is the immediate reason or mechanism that leads to an effect or event. It is often the most visible or direct cause, but it might not be the underlying reason for an issue.
Characteristics: Easily identifiable, directly leads to the outcome, and is often a part of a chain of events.
Root Cause
Definition: The root cause is the fundamental, underlying reason for a problem or event. Addressing the root cause is necessary to prevent the recurrence of the issue.
Characteristics: Often requires deeper analysis to identify, addresses the underlying issues leading to the proximate cause, and solving it usually leads to long-term resolution.
EG 1. Why did I lose my job?
Proximate cause: the company was experiencing financial difficulties and could not continue to pay all its employees.
Root cause: I was not of particular value to the company and they could survive easily without me.
EG 2. Vehicle Accident
Proximate Cause: The car's brakes failed, leading to a crash.
Root Cause: Lack of regular maintenance and inspection of the vehicle's braking system.
What is a logic tree?
You make a logic by finding the high level ‘problem statement’ and then MECEing it into sub components.
If you disagree, there is an assumption in your Logic Tree that is different.
Find the assumption where you differ, ie MECE your way to it, then do the work to figure out what is right / wrong about the assumption.
This is not about data, this is about synthesis of data. This is not about the loudest or most senior voice in the room, it’s about the highest quality decision.
Non business example: Choosing a vacation destination between two friends, Jordan and Taylor.
1. Decide on the problem statement (hypothesis).
Choosing a vacation destination between two friends, Jordan and Taylor.
2. Build and agree on a Logic Tree by MECEing the problem statement.
Problem Statement: Jordan and Taylor cannot agree on a vacation destination.
A. Preferences in Activities
A1. Outdoor vs. Urban Activities
A1a. Hiking, Camping
A1b. Museums, Shopping
A2. Relaxation vs. Adventure
A2a. Spa, Beach
A2b. Skydiving, Surfing
B. Budget Constraints
B1. Accommodation Costs
B1a. Luxury Hotel vs. Budget Stay
B1b. Flight Costs, Car Rental
B2. Travel Expenses
B2a. Deals and Discounts Availability
B2b. Off-peak vs. Peak Season Travel
3. Find out where in the Logic Tree you disagree (eg you view good, they view bad). IE find the root cause of disagreement.
Preferences in Activities: Jordan prefers outdoor and adventure activities like hiking and skydiving, while Taylor leans towards urban exploration and relaxation at a spa or beach.
Budget Constraints: Taylor is concerned about keeping costs low, favoring budget accommodations and off-peak travel, whereas Jordan is more flexible on spending, considering even luxury stays and peak season travel.
Root Cause Identification:
The fundamental disagreement arises from a difference in what each person wants to experience on the vacation (adventure and nature vs. urban exploration and relaxation) and how much they are willing to spend.
4. Do the work to find out which viewpoint best reflects reality at this point.
Solution Development: To find a compromise, they could look for a destination that offers a mix of both worlds: perhaps a coastal city where they can enjoy urban amenities as well as outdoor adventures. They could also explore finding budget-friendly options during the shoulder season, which balances costs and the availability of activities.
If you only take one thing away
Done well, disagreements are a way to learn. Hopefully you can look forward to disagreements!