Art and Science of Problem Solving

Purpose

Ex: My sale is going by 30% - this is not the problem. this is the situation.

If I were given one hour to save the world, I would spend 59 minutes defining the problem and one minute solving it. - Albert Einstein

Breaking down a given problem or a situation into sub parts to understand where does the actual problem lie. It involves deconstructing the problem down to its core elements and building up from there
In simple words getting into the depth of a problem by asking series of questions to figure out where is the root of the problem and then solve for that part of the problem with best possible solution
Most importantly being curious in a genuine way and not to be in hurry to provide a solution without knowing the exact nature of the problem in depth Ex: Elon Musk - journey to mars, building rockets Ex: India's Mangalyaan Mission costed just 11% of Nasa's Maven Mars Mission

By asking right set of questions to divide it into sub parts! As simple as that! ...then keep asking questions till we get all the answers or know what all questions you need to solve to arrive at an overall solution. Ex: Doctor asking questions for diagnosis Ex: Investigators are asking why a particular crime occurred, who will benefit, who is connected.

5 Why's framework - Ask 5 Why's to get to the depth of the problem. One of the most common frameworks for root cause analysis scenario. Developed by Toyota.
- Ex: Why did the robot stop
- Observe the production floorer without preconceptions
Factor Map Creation
- Exhaustive and Exclusive breakdown of the key question
- Before doing EDA, create things like factor maps or you will be limited and biased by the data.
Fishbone framework - Inspecting each of the key areas where to understand where is the problem lying. This is similar to Factor map creation but is more common in finding out issues for an existing process
- Analyzing each part of the processes, services and products to identify the core issue or set of issues leading to the overall problem at hand
- Ex: Why was the coffee bad?

Create your set of hypothesis basis the listed factors
Check data to prove or disprove your hypothesis
If certain data elements/metrics are not available then find proxies before deciding to drop that variable
Exploratory data analysis to understand the nature of relationships between different metrics/variables
Create derived variables to measure impact of variables tin a more holistic way and to reduce number of variables like Change in profit% instead of actual
Find correlation between different variables to understand if certain variables can be dropped
Research about the business you are you are working with to gain required domain knowledge to solve the problem well.
- For instance: In pharma there are regulations around promotional activities. So any recommendations around retail kind of promotions would be null and void for the clients.
Present your analysis in a visual form like by creating a dashboard, using relevant charts, having a proper flow of analysis and eventually providing findings, insights and recommendations