7-Steps-Problem‐Solving-Framework

Step-1: Define The Problem

  1. Rushing to find a solution reduces chances of finding good solution

  2. Identify the knowns, unknowns and challenges

  3. Identify if good-enough outcome works or do we want perfect solution

  4. List any known assumptions

  5. List verifiable facts

  6. Understand the Problem

  7. Understand the context

  • Identify the stakeholders

  • Identify the constraints

  • Identify the non-negotiables

  • Identify the success criteria

  • Identify the desired accuracy level [perfect response or good-enough response]

  1. Eval against the smarter framework

  • Is problem specific

  • is it measurable

  • is it actionable

  • is it relevant

  • is it timebound

  • is it ethical review

  • is there reputational risks

  • is there material risks

  • is there discriminational risks

  • is there exploitational risks

Step-2: Disaggregate/Breakdown the Problem

Step-1: Build Your Logical Tree

  1. Pinpoint, levers, factors, and components associated with problems

  • Trunk = Problem

  • Big branches = factors

  • Small Branches = levers

  • Twigs = components

  1. List everything that contributes to problem

  2. Use Stickies

  3. Brainstorm with Team

  4. Logical trees can be worked either of following two ways:

  • Deductively aka from problem to cause

  • Inductively aka from cause to problem

  1. Details evolve our logical tree into hypothesis tree

  2. Logical trees should be mutually exclusive aka there should be minimal to no overlap and collectively exhaustive aka include all ways to look at the problem

Step-2: Evolve Logical Tree into Hypothesis Tree

  1. This is one of the many probable cause that needs to be validated

  2. Specific hypothesis are better for testing

  3. Easy to test hypothesis allows us to look at problem from different perspective

Step-3: Use Cleaving Frames

Use different perspective to evaluate the how long it will take to solve the problem

  1. work/play frame suggests to consider how much effort and for how long we have to apply at/towards something to meet our end goal

  2. supply/demand frame suggests to consider how much of anything is needed to meet the demand of that item

  3. near/long frame suggests to consider short and long term implications of our decisions before making any decisions

  4. cost/benefit frame suggests to consider how much something is costing at hourly or financial basis and if the benefit yielded is signifcantly higher than that

Step-3: Prioritization of Problem

Find your critical path

  1. Direct effort with aim to apply least effort for max returns

  2. Collaborate and avoid bias

  3. Neglecting prioritization could lead to delusion of focus

  4. Optimize your critical path

  • Tackle things that have high impact and your ability to influence the actions necessary for outcomes is high

  • Identify and eliminate things that offer little help toward solution

  • Identify and eliminate things that are too hard to tackle

  1. 80% of outcomes are result of 20% of effort or 20% causes are reasons for 80% of problems

  2. Identify

  • Who does what

  • When it is due

  • Why it is needed

  • What does end-product look like

Step-4: Work Planning

For Multi Month Projects

  • Use Chunky Work Plan

  • Prepare for 2 weeks worth of work ahead

  • Regroup

  • Analyse

  • Ask, are task for the week complete,

  • if yes: move forward

  • if no: start steps 1-3

For One Day Answers

We do all of the things state under "Multi Month Projects Work Plan" in one day by doing followin

  • Identify Situation

  • Record Observations

  • Understand and Share Implications

Note: Combining chunky work plan and gantt chart leads to balanced work plan

  • Use detailed work planning with due dates for individual tasks

  • Ensure delivery of individual work products is well within total allocated time for whole project

  • Leverage those with expertise you lack

  • Adopt a Proactive outlook

  • Diversity of perspective will enrich problem solving

  • Even with hammer in hand, not everything is a nail

  • Speak last to avoid sunflower effect, aka, its when your followers follow your decision because their view might contradict yours

Biases in Decision Making

  1. Availability Bias suggests:

  • We rely on what technique or method worked most recently

  • ignore the unique characteristics of problem [how new problem differs from last similar encounter]

  • Did we take into account everyone's perspective to consider effect before making a decision

  1. Anchoring Bias suggests:

fixating on stats over present day situations

  1. Confirmation Bias suggests:

  • We favor the information that supports our hypothesis against the one that is contrary to our hypothesis

  1. Loss-aversion Bias suggests:

  • Prefer avoiding loss over acquiring gains even when the possibility of postitive and negative outcomes are equivalent

  1. Over-optimism Bias suggests

  • overly positive outlook

  • overestimate likelihood of positive outcomes

  • unrealistic expectations

  1. Democratic Bias

  • It occurs when people avoid sharing their opinion to avoid dominance bias

Step-5: Analysis of Problem

  1. Get a 30,000 foot view of the problem

Heuristics System suggests

  • Solution with least assumptions tends to be correct one

  • Evaluate if 80% of problem stems from 20% of causes

  • Determine Breakeven point

  • to calculate sales volume needed to cover fix costs and variable costs per unit sold

  • Evaluate against rule of 72 for financial scenarios to determine no years required to double the invested money

  • 72/(Annual growth rate of invested money)

Pitfalls of Heuristics

  1. Substitution Heuristics - we answer question other than what was asked

  2. Availability Heuristics - we overestimate probability of occurrence of something because it is easy to remember

Summary of Statistics suggests

  1. Calculate average, median and mode to measure central tendency and data distribution

  2. Analyze Outliers

  • To determine best and worst performers

  • To reveal error in data sampling

  • To reveal optimal and sub-optimal practices

Questions Based Method suggests:

Systematically ask why, 5 times consecutively

  • To help drill down root cause of problem

  • To help shift focus to most critical aspects

Driver of the Situation suggests:

  1. Determine the data quality

  2. Ensure data is reliable

  3. For complex data, use multi-variance

  4. Use A/B Testing to generate data when data is lacking

  5. Bayesian is used to make inference from imperfect data

Prediction

  1. Account change in behavior of actor

  • Use ML or scenario modelling

  1. Behavior prediction of rival is through game theory

Step-6: Synthesis of Problem-Solution Narrative

  1. Convert Insights to story

  2. Work in teams

  3. Discuss with teams to find relevant roles and responsibilities

  4. Work on central idea/thesis/governing thought

  • Define solution in one sentence

  • Define recommendation to achieve the solution

  1. Use either top-down [Hypothesis to Solution] or bottom-up [Solution to Hypothesis]

  2. Cycle between above two approaches as needed

  3. Keep in mind

  • Constantly review problem statement and understanding while considering newly acquired information

  • Remember to remove any irrelevant facts and information

  • Consider only verified facts for conclusion

  1. Ask yourself

  • Did you understand the problem correctly

  • Do you understand the constraints applied

  • Do you understand all the potential or known challenges you will encounter

  • Do you have clear set of success criteria

  • Are we ready to share solution with stakeholder

  • Get second set of reviews before sharing your conclusions

Step-7: Communication of Problem-Solution

  1. Create persuasive narratives

  2. Communicate via compelling storytelling

  3. Explain, why change is necessary

  4. Organize thoughts

  • from summary to reasoning

  • start with governing thought - > state desired action - > include supporting argument with data

Frameworks for Crafting Frameworks

  1. Argument Structure suggests

  • suggesting change is introduced first

  • causes with proof

  • solution with data

  1. Grouping Structure suggests

  • add Recommendations to do x/y/z

  • add reasons

  • add data

  1. Reveal Structure suggests

  • What to do when audience skeptical from beginning

  • Ask questions again and again until responses lead to conclusion

Points to Remember

  • Use whiteboards and sticky notes

  • Use props and demos

  • People love stories and visuals

  • People consumer information better with visuals

  • Visuals are needed for memorable content

  • let of quest for perfect response

  • flexibility and iteration are key

Last updated