Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

CONSULTANT SKILLS SERIES · COURSE 1

Structured Problem Solving

& Hypothesis-Driven Thinking
The foundational consulting skill — from problem to recommendation

8 40+ 6 100%
Modules Examples Frameworks Practical

HARTS Consulting · Internal Use Only Page 1

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

Table of Contents

Module 1 The Consulting Problem-Solving Mindset........................................................3

Module 2 Issue Trees and MECE Structuring.................................................................6
Module 3 Hypothesis-Driven Thinking...........................................................................10
Module 4 The Problem-Solving Process End-to-End....................................................14
Module 5 Data Gathering and Analysis.........................................................................18
Module 6 Synthesis — From Findings to Insights.........................................................22
Module 7 Building the Recommendation.......................................................................26
Module 8 Communicating Your Answer........................................................................30

HARTS Consulting · Internal Use Only Page 2

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 1

The Consulting Problem-Solving Mindset

Why consultants think differently — and how to develop it

Before any framework, any tool, any slide deck — consulting starts with a mindset. The way a
consultant is trained to think about problems is fundamentally different from how most
professionals approach them. This module builds that mental foundation.

The Core Difference: Answer-First Thinking

Most professionals are trained to gather data, analyze it thoroughly, and then form a conclusion.
Consultants are trained to do the opposite: form a hypothesis first, then gather only the data
needed to prove or disprove it. This is not cutting corners — it is ruthless efficiency under time
pressure.

Traditional Thinking Consulting Thinking

Gather all data → Analyze → Conclude. Form hypothesis → Test with targeted data →
Thorough but slow. Often leads to analysis Refine or reject. Faster. Directed. The analyst is
paralysis. The conclusion can be a surprise even never surprised because they started with a bet.
to the analyst.

EXAMPLE — MCKINSEY DAY 1 VS DAY 30

A McKinsey team studying a retailer's declining margins doesn't spend 4 weeks gathering
data before forming a view.
On Day 1 they ask: 'Is this a revenue problem or a cost problem?' They form a hypothesis
('likely cost — COGS have risen in 3 of last 4 quarters')
and spend the next week gathering only data that tests that hypothesis.
By Day 7 they either confirm the hypothesis and go deeper, or reject it and pivot.
This is why consulting projects that take 6 months produce insights that internal teams
couldn't generate in 3 years.

The Five Consulting Mindset Principles

1 Always start with the problem, not the solution

The most common mistake freshers make: jumping to solutions before the problem is well-
defined.
A client who says 'we need a new CRM' may actually have a sales process problem, a data

HARTS Consulting · Internal Use Only Page 3

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

quality problem,
or a team capability problem — and a new CRM won't fix any of them.
Your first job is always to define and validate the actual problem before proposing anything.

2 Be comfortable with ambiguity — but don't stay in it

Real consulting problems don't come with clean data and clear answers. You will often have
to make a recommendation
with 60–70% of the information you'd ideally want. The skill is knowing when you have
enough to be directionally right
and moving — not waiting for certainty that will never arrive.

3 Separate fact from inference from assumption

Rigorous thinking requires labeling what you know vs. what you're inferring vs. what you're
assuming.
Fact: Revenue declined 12% last quarter. Inference: This is likely driven by price erosion in
the SME segment.
Assumption: The SME segment accounts for 40% of revenue. The moment you confuse
these three, your analysis becomes fragile.

4 Think in structures, not lists

A list is a collection of items. A structure is a set of items organized by a logic that reveals
relationships.
'We have problems with quality, delivery, and customer service' is a list.
'Our customer satisfaction gap is driven by failures across the product delivery value chain
— from manufacturing quality through last-mile logistics to after-sales support' is a structure.
Structures create clarity. Lists create noise.

5 Kill your darlings

The worst thing that can happen to a hypothesis is that you fall in love with it.
The moment you start protecting your hypothesis from contrary evidence, you've stopped
being an analyst
and started being an advocate. The best consultants actively look for evidence that they're
wrong.
Being proven wrong early is far less expensive than being proven wrong in front of a client.

The Fresher's Most Common Mindset Error

Confusing activity for progress. Filling slides, gathering data, running analyses — without
stepping back to ask: 'What question am I trying to answer right now, and does this activity
answer it?'

HARTS Consulting · Internal Use Only Page 4

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

Every 2 hours of analysis should be preceded by 10 minutes of structured thinking about what
you're actually trying to learn.

HARTS CONTEXT — CLIENT DIAGNOSTIC WORK

When scoping a new engagement for a client like GMR, the first instinct is to start gathering
data — org charts, financials, process maps.
The consulting approach: spend the first 48 hours forming a hypothesis about what the core
problem is likely to be,
based on the engagement brief, the client's sector, and analogous situations from prior work.
That hypothesis shapes the diagnostic — and saves 2 weeks of unfocused data collection.

HARTS Consulting · Internal Use Only Page 5

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 2

Issue Trees and MECE Structuring

The backbone of every consulting analysis

The issue tree is the single most important tool in the consultant's toolkit. It is the practice of
decomposing a complex problem into its component parts in a structured, logical, exhaustive
way. Mastering issue trees is what separates a consultant who can handle any problem from
one who can only handle familiar ones.

What is MECE?
MECE: Mutually Exclusive, Collectively Exhaustive
Mutually Exclusive: No overlap between categories. Each item belongs in exactly one bucket.
Collectively Exhaustive: No gaps. Together, the categories cover the entire problem space.

MECE is not just a formatting rule — it is a test of the quality of your thinking.
If your categories overlap, you're double-counting. If there are gaps, you may be missing the
answer.

MECE IN PRACTICE — DIAGNOSING REVENUE DECLINE

Non-MECE: 'Revenue is down because of pricing issues, competitive pressure, and losing
customers.' These overlap — competitive pressure causes pricing issues and customer loss.

MECE: Revenue = Volume × Price. Volume is down (fewer units sold). OR Price is down
(same units at lower price). OR both.
Volume = New customer acquisition × Retention. Price = List price × Discount rate × Mix
shift.
Each branch is exclusive. Together they cover 100% of the revenue equation.

Types of Issue Trees

Type When to Use Example Structure

Logic Tree When you have a hypothesis to 'Profitability is low' → Is it a revenue

(Hypothesis) test. Work top-down from answer to problem? → Is it a cost problem? →
evidence. Is it a capital efficiency problem?

Issue Tree When you don't know the answer 'Why is customer satisfaction low?'

HARTS Consulting · Internal Use Only Page 6

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

(Diagnostic) yet. Work from problem to possible → Product quality → Delivery

causes. experience → After-sales service →
Price-value perception

How Tree When you need to generate options 'How do we increase revenue by
(Solution) for solving a defined problem. 20%?' → Grow existing customers
→ Acquire new customers → Enter
new markets → Increase pricing

Building an Issue Tree — Step by Step

1 State the problem as a question

Start with a crisp, answerable question. Not 'our margins are declining' (observation) but
'Why have EBITDA margins declined from 18% to 11% over the past three years?'
(question).
A well-formed question already constrains the answer space.

2 Identify the first level of branches

Find the 2-4 major categories that are MECE at this level.
Test: 'Can the answer to our question live in exactly one of these branches?' If yes, you're
MECE.
Common first-level structures: Revenue/Cost, Internal/External, Short-term/Long-term,
People/Process/Technology.

3 Decompose each branch one level deeper

Each branch should break into 2-4 sub-branches that are MECE within that branch.
Stop decomposing when you reach something that can be directly measured or tested.
Typical trees have 3-4 levels. More than 4 levels suggests you're getting lost in the details.

4 Prioritize branches before analyzing

Not all branches matter equally. Before going deep on analysis, make an initial judgment:
Which branches are most likely to contain the answer? Which have the most potential
impact?
This is where hypothesis thinking enters — you form a view on where to look hardest.

5 Test, not prove

The goal is to determine which branches are true — not to confirm what you already
believe.
For each prioritized branch: identify the 1-2 pieces of data that would tell you if this branch is

HARTS Consulting · Internal Use Only Page 7

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

the answer.
Gather that data. Make a call. Move on.

FULL EXAMPLE — GMR AIRPORT CONCESSION REVENUE

Problem: 'Why is non-aeronautical revenue per passenger at the terminal 30% below
benchmark?'

Level 1 (MECE): Volume of spend per passenger | Categories of spend offered

Volume branch: Dwell time per passenger | Conversion rate (entering vs. buying) |
Average transaction value
Categories branch: F&B offering | Retail offering | Services (lounges, parking, WiFi)

Initial hypothesis: Conversion rate is likely the culprit — not dwell time or category breadth.
Test: Pull conversion data by terminal zone and by concession category.
Finding: Conversion is 18% vs. 31% benchmark — but only in the pre-security zone. Post-
security is comparable.
Diagnosis: The problem is pre-security layout and wayfinding, not the concession mix itself.

Common MECE Failures and How to Fix Them

Failure Example Fix

Overlap between branches Revenue issues AND pricing Ask: Can something belong to
issues (pricing is part of two branches at once? If yes,
revenue) restructure.

Missing a whole category People and Process — Ask: If none of these branches
missing Technology are the answer, where else
could the answer live?

Too many branches at one 8 branches at Level 1 Group into meta-categories. 8

level items is a list. 3-4 items is a
structure.

Branches at different levels Revenue, Customer Ensure all branches at a given

of abstraction Satisfaction, Pricing, NPS level are the same type of
thing.

HARTS Consulting · Internal Use Only Page 8

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 3

Hypothesis-Driven Thinking
The art of starting with the answer

Hypothesis-driven thinking is the most powerful — and most misunderstood — skill in

consulting. It is not guessing. It is informed, structured speculation backed by logic, prior
experience, and early data. It transforms the problem-solving process from an open-ended
search into a directed investigation.

What a Hypothesis Is — and Isn't

A Hypothesis IS A Hypothesis IS NOT

A specific, testable, falsifiable statement about A vague directional guess. 'The problem might
the answer to a problem. It makes a claim that be in operations' is not a hypothesis. It lacks
can be proven true or false with evidence. specificity and cannot be directly tested.

A Hypothesis IS A Hypothesis IS NOT

An informed starting point that shapes your A commitment. You must be willing to abandon it
analysis — not a conclusion you're defending. the moment the evidence says so.

Anatomy of a Good Hypothesis

The Hypothesis Formula

We believe that [specific claim about the situation] because [logical reasoning / early evidence].
If this is true, we would expect to see [observable test condition].
We can test this by [specific data to gather or analysis to run].

HYPOTHESIS IN PRACTICE — RETAIL CHAIN PROFITABILITY

Problem: A mid-sized retail chain's net margin has fallen from 7% to 3% in 18 months.

Hypothesis: 'We believe the margin decline is driven primarily by increased logistics and last-
mile costs,
not by pricing pressure or sales volume, because the shift coincides with the company's
expansion into Tier 2 cities
where cost-per-delivery is significantly higher than the metro baseline.'

HARTS Consulting · Internal Use Only Page 9

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

'If this is true, we would expect to see: Gross margin broadly stable, COGS relatively flat,
but operating costs — specifically freight and warehousing — up as a % of revenue.'

'We can test this by: Pulling a cost waterfall by cost category for the last 6 quarters,
segmented by metro vs. Tier 2 operations.'

The Hypothesis Pyramid

A well-formed hypothesis is never a single statement — it is a pyramid of linked hypotheses,
each one supporting the one above it.

Governing The top-line answer to the problem. 'The company should exit the SME
Hypothesis segment and focus resources on enterprise accounts.'

Supporting The 3-4 reasons the governing hypothesis is true. Each one supports the
Hypotheses overall claim.
1. SME accounts generate 30% of revenue but 60% of support costs.
2. Enterprise accounts have 3× higher lifetime value.
3. The sales team's skills are better matched to complex enterprise selling.

Evidence For each supporting hypothesis, the specific data points needed to confirm it.
Hypotheses
'SME accounts generate 60% of support costs' → Test: Pull support ticket
volume and resolution time by account tier.

How to Form a Strong Initial Hypothesis

1 Use analogues from similar situations

The fastest route to a good hypothesis is asking: 'Have I seen this before?' or 'Has
someone seen this before?'
If a logistics company's margins are falling, look at how similar cases resolved in the
industry.
Pattern recognition is not cheating — it is one of the primary reasons clients hire
consultants.

2 Apply first principles

Break the situation down to its fundamentals. For any business problem, the answer must
ultimately
live in one of a small number of value drivers: revenue, cost, capital efficiency, or risk.
First principles thinking narrows the hypothesis space before you've seen a single data

HARTS Consulting · Internal Use Only Page 10

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

point.

3 Look at what's changed

Most business problems are caused by something that changed. Revenue was fine — then
it wasn't.
Margins held — then they didn't. Your hypothesis should almost always reference a change:
'We believe X because Y changed at approximately the same time the problem emerged.'

4 Prioritize with the 80/20 rule

Not all hypotheses deserve equal testing time. Apply Pareto: which hypothesis, if true,
would explain 80% of the problem? Start there. Most problems have one or two root causes
that account for the vast majority of the impact. Hunt for those first.

HYPOTHESIS-DRIVEN VS. DATA-DRIVEN — SIDE BY SIDE

Data-driven approach: Pull all available data across revenue, cost, customer, operations,
and HR.
Spend 3 weeks cleaning and analyzing. Present findings. Total time: 6 weeks.

Hypothesis-driven approach: Spend 2 hours forming 3 hypotheses. Identify the 3 datasets

that would test them.
Spend 1 week gathering and analyzing only those datasets. Confirm or reject each
hypothesis.
Form a refined second-generation hypothesis. Test. Present. Total time: 2 weeks.

Same answer. 3× faster. This is why top consulting firms charge what they charge.

HARTS Consulting · Internal Use Only Page 11

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 4

The Problem-Solving Process End-to-End

From ambiguous brief to clear recommendation

This module integrates everything from Modules 1-3 into a complete, repeatable problem-
solving process. This is the process that top consulting firms use — from a startup strategy
engagement to a global transformation program. The scale changes. The process doesn't.

The Six-Stage Consulting Problem-Solving Process

1 Problem Definition — What are we actually solving?

Duration: 10-20% of total project time. Often underinvested and almost always worth more
time than it gets.

Outputs: A problem statement that is specific, bounded, and answerable.

A set of clarifying questions that surface hidden assumptions.
Agreement with the client on what a good answer looks like.

The classic mistake: starting analysis before the problem is agreed. You can spend 3 weeks
analyzing the wrong question.

Tool: The Problem Definition Document — one page that states the problem, the context,
the constraints, the success criteria, and what is explicitly out of scope.

2 Structure — How do we break the problem down?

Duration: 5-10% of total time. Should happen before any data is gathered.

Outputs: An issue tree that is MECE. A prioritized set of sub-questions to answer.

An initial hypothesis about where the answer is likely to live.

The classic mistake: building the structure in your head rather than on paper.
Invisible structures cannot be challenged, refined, or shared.

Tool: Issue tree (Module 2). Initial hypothesis statement (Module 3).

3 Analysis Planning — What do we need to prove or disprove?

Duration: 5% of total time. This step is almost always skipped by freshers — and it shows.

HARTS Consulting · Internal Use Only Page 12

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

Outputs: An analysis plan that maps each hypothesis branch to:

→ The specific question it answers
→ The data needed
→ The analysis method
→ The output format (chart, table, model)
→ The person responsible

This step prevents the most common consulting waste: collecting data you never use.

4 Data Gathering — What do we actually know?

Duration: 30-40% of total time.

The key discipline: gather only what your analysis plan calls for. Resist the urge to gather
everything.
When you find unexpected data, update your hypothesis — don't just add it to the pile.

Sources: Interviews (primary), internal data (secondary), market data (secondary),

benchmarks, observations, analogues from other industries.

Quality over quantity: One credible data point that tests a hypothesis is worth more than
100 data points that don't connect to the core question.

5 Synthesis — What does it all mean?

Duration: 20-30% of total time. The hardest step and the one that most clearly separates
junior from senior consultants.

Synthesis is not summarizing data. It is extracting the 'so what' — the insight that the data
reveals when you look at it through the lens of the problem.

Output: A set of key findings, each expressed as an insight (not a data point),
linked back to the hypothesis structure, with a clear implication for the recommendation.

6 Recommendation — What should they do?

Duration: 10-15% of total time. Paradoxically, the output that takes the least time to
generate
if the preceding 5 steps are done well.

A good recommendation has: A clear answer (what to do), a rationale (why), a risk
assessment
(what could go wrong), an implementation path (how), and success metrics (how we'll know

HARTS Consulting · Internal Use Only Page 13

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

it worked).

The golden rule: Never let the recommendation be a surprise. By the time you present the
final answer,
the client should have been involved enough in the journey that they co-own the conclusion.

FULL PROCESS — STRATEGY ENGAGEMENT AT A LOGISTICS COMPANY

Stage 1: Problem defined as: 'Should we expand into cold-chain logistics, and if so, how?'
Stage 2: Issue tree built around: Market opportunity | Competitive position | Capability to
deliver | Financial viability
Stage 3: Hypothesis prioritized toward capability gap (prior experience suggested this is
where logistics cos. fail most)
Stage 4: Interviewed 12 cold-chain customers, analyzed 3 competitor P&Ls, ran a gap
analysis against internal capabilities
Stage 5: Synthesis — market is attractive, competitors are weak on technology, but our
client lacks refrigerated fleet and cold-storage network
Stage 6: Recommendation — enter via acquisition of a regional player, not organic build. 18-
month integration roadmap with 3 milestones.

The Tempo of Problem Solving

One of the hardest things for freshers to internalize: the process is not linear.
New evidence will cause you to revise your hypothesis and return to structuring mid-analysis.
A client conversation will redefine the problem in week 3 and require restructuring the entire
approach.
This is not failure — this is how real consulting works.
The process is a scaffold, not a cage. Use it to organize your thinking, not to trap it.

HARTS Consulting · Internal Use Only Page 14

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 5

Data Gathering and Analysis

Gathering only what you need — and analyzing it with discipline

Data analysis in consulting is not about doing the most sophisticated analysis — it is about
doing the right analysis. The most powerful analyses are often the simplest ones, applied to
exactly the right question.

The Analysis Planning Matrix

Before touching a single spreadsheet, complete this matrix for every sub-question in your issue
tree:

Sub-Question Data Needed Analysis Type Expected Output

Is this a revenue or P&L for last 6 Trend analysis, Revenue vs. cost
cost problem? quarters, by waterfall chart contribution to margin
business unit delta

Which customer Churn data by Segmentation, Ranked segment risk

segment is most at segment, NPS by correlation table with revenue at
risk? segment, revenue stake
by segment

Is our pricing Internal price list, Benchmarking, Price position map vs. 3
competitive? competitor pricing price-volume key competitors
(public), win/loss analysis
data

Primary vs. Secondary Research

Primary Research Secondary Research

Data you generate yourself: interviews, surveys, Data that already exists: internal reports, industry
observation, workshops. Highest reliability for databases, public filings, benchmarks. Faster.
opinions, motivations, and qualitative insights. Always start here before conducting primary
Time-intensive — plan carefully. research. Risk: may be outdated, out of context,
or low quality.

HARTS Consulting · Internal Use Only Page 15

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

The Consulting Interview — Getting Insight, Not Just Information

Interviews are the primary research tool of choice in consulting. A well-conducted 45-minute
interview can replace weeks of secondary research. A poorly conducted one wastes everyone's
time and damages client relationships.

1 Prepare with a hypothesis, not just questions

Don't walk into an interview with a list of generic questions. Walk in with a hypothesis you're
trying to test.
'I believe the operational delays are caused by the handoff between procurement and
operations — I want to test that.'
This changes what you ask, how you listen, and what you follow up on.

2 Use the funnel structure

Start broad (context, role, general observations), then narrow (specific incidents, examples,
data),
then deep (implications, root causes, recommendations).
The richest insights usually come in the last 15 minutes — after the interviewee is
comfortable and talking freely.

3 Listen for the unspoken

What is the interviewee not saying? What do they hedge on? What question makes them
pause?
Silence is often the most valuable data. When someone pauses before answering, follow
up:
'That seemed like a complex question — what's the nuance there?'

4 Always end with 'who else should I speak to?'

The best interviewees refer you to the people they respect — which is often exactly who you
need to reach.
This question also signals to the interviewee that you're building a complete picture, not
cherry-picking.

The Most Useful Analytical Techniques

Technique When to Use Consulting Example

Trend Analysis When you want to understand Revenue per customer declining
how something has changed over over 6 quarters — is it
time acceleration, plateau, or

HARTS Consulting · Internal Use Only Page 16

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

reversal?

Waterfall / Bridge When you want to explain what 'What drove the Rs. 40Cr profit
Chart drove a change between two drop?' — break it into volume,
states price, mix, and cost components

80/20 / Pareto When you want to identify the Top 20% of SKUs drive 80% of
Analysis vital few from the trivial many margin — focus the cost
reduction effort there

Benchmarking When you need to establish Cost-per-transaction at 2.3% vs.

whether a metric is good or bad industry benchmark of 1.1% —
confirms a cost problem

Root Cause Analysis When you want to get past 'Deliveries are late' → why? →
(5 Whys) symptoms to causes 'Trucks dispatched late' → why?
→ 'Driver scheduling gaps' →
why?...

Scenario / Sensitivity When the recommendation 'At what revenue growth rate
depends on uncertain does Option A outperform Option
assumptions B?' — find the crossover point

ANALYTICAL DISCIPLINE — THE MINIMUM SUFFICIENT ANALYSIS PRINCIPLE

A common mistake: running every analysis available and presenting all of it.
The result: slides full of charts, none of which clearly answers the core question.

The discipline: for each analysis you run, ask 'Does this directly answer a sub-question in my
issue tree?'
If no → don't run it. If yes → run it, extract the insight, and present only the insight (not the
raw output).

Rule of thumb: for every 10 analyses a consulting team runs, 3 make it into the final
presentation.
The other 7 are how you figured out which 3 mattered. This is normal and expected.

HARTS Consulting · Internal Use Only Page 17

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 6

Synthesis — From Findings to Insights

The hardest skill in consulting

Synthesis is the transformation of raw findings — data points, interview quotes, analytical
outputs — into insights that directly answer the client's problem. It is what separates a junior
analyst from a consultant, and a consultant from a partner. It cannot be automated, delegated to
data, or replaced by a framework.

Finding vs. Insight — The Critical Distinction

Finding (raw) Insight (synthesized)

'Revenue has declined 12% over 3 years. Gross 'Our profitability problem is entirely cost-driven —
margin has held. Operating costs have increased not a revenue weakness. Operating cost growth
24%.' This is what the data says. It describes. It at 24% in 3 years is unsustainable and points to
does not conclude. a structural inefficiency, not a cyclical fluctuation.'
This is what the data means.

The 'So What' Test

After every finding, ask: 'So what?' Three times.
Finding: 'Operating costs increased 24%.' So what?
'The margin decline is cost-driven, not revenue-driven.' So what?
'The intervention should focus on cost reduction, not revenue growth.' So what?
'Recommending a pricing initiative or sales investment would be the wrong answer for this
client.'

The third 'So what' is usually where the insight lives.

The Synthesis Process

1 Assemble all findings in one place

Before synthesis, collect everything: analysis outputs, interview notes, benchmark
comparisons, observations.
Use a synthesis wall (physical or virtual) — post every finding as a separate note.
No editing at this stage. You want everything visible before you start organizing.

HARTS Consulting · Internal Use Only Page 18

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

2 Group by theme, not by source

The mistake: organizing findings by where they came from (interview findings, financial
findings, market findings).
The right approach: group by what they mean — what pattern or theme do they point to?
A finding from an interview and a finding from the financial model that both point to the
same root cause belong together.

3 Name each cluster with an insight, not a label

'Cost Issues' is a label. 'Logistics costs are 40% above benchmark and are the single
largest driver of margin decline' is an insight.
Every cluster must be named with a statement that makes a claim about what is true.
If you can't name it with a claim, you haven't finished synthesizing it.

4 Test the insights against the original problem

Do your insights, taken together, answer the original problem statement?
Are they MECE — do they cover the full problem space without overlapping?
Are they mutually reinforcing — do they build a coherent story, or do they contradict each
other?
The synthesis is complete when the answer to the original question is clear and defensible.

5 Build the logic chain

Connect your insights into a chain of reasoning:
Because of X → we know that Y → which means Z → therefore we recommend W.
If any link in this chain is weak (missing evidence, assumption-dependent), flag it before the
client does.
A chain with one weak link is weaker than you think — clients pull hard on weak links.

SYNTHESIS IN ACTION — ERP IMPLEMENTATION HEALTH CHECK

Raw findings from a 4-week diagnostic of an SAP implementation:
→ 34% of master data records have errors (data quality analysis)
→ Finance team has not attended 3 of 5 design workshops (attendance log)
→ 11 of 18 KDDs (Key Design Decisions) are unresolved past their due date (project
tracker)
→ Business sponsor attended only the kickoff — not since (interview with PM)
→ Integration testing schedule is compressed by 3 weeks due to early delays (plan review)

Unsynthesized: 'There are data quality issues, attendance problems, and schedule risks.'

Synthesized: 'The implementation is at high risk of go-live failure. The core issue is not
technical — it is

HARTS Consulting · Internal Use Only Page 19

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

governance. Business ownership is absent at exactly the points where it is most needed:
design decisions, data cleansing, and testing. Without executive re-engagement in the next
2 weeks,
we recommend a 6-week go-live delay to avoid a failed cutover.'

Triangulation — Making Insights Defensible

A single data point is a coincidence. Two is a pattern. Three is a finding. The discipline of
triangulation — finding at least three independent sources of evidence for each key insight — is
what makes your synthesis defensible under pressure from a skeptical client.
• Financial data says costs are up → confirms direction
• Benchmarking says costs are above industry → confirms materiality
• Operations interview says 'we've known about this for 18 months' → confirms root cause
is known, not hidden

Together these three make an insight that holds under scrutiny. Any single one of them could be
explained away. Together, they cannot.

HARTS Consulting · Internal Use Only Page 20

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 7

Building the Recommendation

From insight to action — the moment of truth

Everything in the problem-solving process builds toward one moment: making a

recommendation. This is the moment a client pays for. A technically brilliant analysis that
produces a vague or hedged recommendation is a failure. A clear, specific, defensible
recommendation is the product.

What Makes a Recommendation Consulting-Grade?

The Five Tests of a Strong Recommendation

1. Specific: Not 'improve operations' but 'reduce procurement cycle time from 18 to 7 days by
centralizing vendor approval authority.'
2. Grounded: Every element of the recommendation traces back to a finding. No
recommendations that appear from nowhere.
3. Actionable: The client can start acting on Monday morning. Vague strategic principles are not
recommendations.
4. Prioritized: When there are multiple recommendations, they are ranked by impact and
sequenced by dependency.
5. Honest about risk: A recommendation without a risk assessment is wishful thinking, not
advice.

The Recommendation Architecture

1 The Governing Recommendation

One sentence that captures the top-line answer to the problem.
This is the answer a CEO should be able to quote after reading the first slide.
'We recommend exiting the B2C segment and redirecting capital to B2B enterprise, which
offers 3× the margin at lower churn.'
If you cannot say your recommendation in one sentence, it is not yet a recommendation —
it is still analysis.

2 Supporting Arguments
The 3-4 reasons the governing recommendation is the right answer.
Each argument corresponds to a key insight from your synthesis.

HARTS Consulting · Internal Use Only Page 21

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

Each insight is supported by evidence from your analysis.

This is the pyramid structure — governing recommendation at the top, arguments in the
middle, evidence at the base.

3 Implementation Path
A recommendation without an implementation path is academic.
Specify: What happens in the first 30 days? The first 90 days? The first 12 months?
Who owns each workstream? What are the major decision gates?
What resources are required (people, budget, technology)?
A rough roadmap that shows the client the path forward is often more valuable than the
recommendation itself.

4 Risk Assessment
Anticipate the top 3 objections before the client raises them.
For each risk: What is the risk? How likely is it? What is the mitigation?
Raising risks yourself signals intellectual honesty. Being surprised by a risk the client raises
is a credibility hit.
Format: Risk table with 3 columns — Risk | Likelihood & Impact | Mitigation

5 Success Metrics
How will we know if this recommendation worked?
Specify 2-3 KPIs that will indicate success, the baseline, the target, and the timeline.
'Customer acquisition cost will reduce from Rs. 2,400 to Rs. 1,600 within 9 months of
implementing the new channel strategy.'
Metrics close the loop between the recommendation and the business case.

Handling Pushback on Your Recommendation

'The data doesn't support this' 'We tried this before and it failed'
Welcome it. Go back to your evidence. Either Ask what was different. Context matters. If the
defend it with specific data points or prior failure is relevant, it should be in your risk
acknowledge the gap and revise. Never bluff. section. If it's not, explain why.

'This is too risky' 'Can we do all of them?'

Separate risk tolerance from risk reality. Quantify Yes, is not always the right answer. Prioritize.
the risk. Then quantify the risk of doing nothing. 'We recommend starting with A because it funds
Often the status quo is riskier. B. Doing all three simultaneously fragments
resources.'

HARTS Consulting · Internal Use Only Page 22

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

RECOMMENDATION QUALITY — BEFORE AND AFTER

BEFORE (typical fresher output):
'We recommend that the company improve its supply chain efficiency through better
processes,
technology investment, and enhanced vendor management.'

AFTER (consulting-grade):
'We recommend a three-part supply chain transformation over 18 months:
1. Consolidate from 14 to 6 tier-1 vendors (Q1-Q2): saves Rs. 8.2Cr in procurement
overhead
2. Implement demand-sensing at 5 high-velocity warehouses (Q2-Q3): reduces stockouts by
est. 35%
3. Deploy vendor scorecarding with quarterly reviews (Q3 onwards): drives 15% on-time
improvement
Combined impact: margin improvement of 2.1 percentage points, ROI positive by month 14.'

HARTS Consulting · Internal Use Only Page 23

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

MODULE 8

Communicating Your Answer

The Pyramid Principle and the art of structured thinking on paper

In consulting, how you communicate your answer is almost as important as the answer itself. A
correct analysis communicated poorly will lose to a slightly weaker analysis communicated
brilliantly. This module covers the frameworks that define how consultants write, structure
slides, and present recommendations.

The Pyramid Principle — Barbara Minto

Developed at McKinsey in the 1970s, the Pyramid Principle is the foundational framework for all
consulting communication — memos, slides, emails, and verbal presentations.

The Core Principle

Lead with the answer. Support it with arguments. Support each argument with evidence.

In academic writing, you build up to the conclusion. In consulting communication, you start with
the conclusion
and spend the rest of the document proving it. The reader always knows where you're going.
They can stop at any level of the pyramid and have a complete, coherent message.

The Pyramid in Practice — Slide Messaging

1 The Governing Thought (Slide Title)

Every slide has one message — and that message is stated in the title as a complete
sentence.
Not 'Revenue Analysis' (a label) but 'Revenue decline is concentrated in the SME segment
and accelerating' (a message).
The governing thought test: if a senior leader reads only the slide titles of your deck,
do they get the full story? In a good consulting deck, the answer is always yes.

2 The Supporting Evidence (Slide Body)

The body of the slide — charts, tables, text — exists solely to prove the title is true.
If an element of the slide body doesn't directly support the title message, remove it.
The most common fresher error: slide bodies that contain interesting data that doesn't

HARTS Consulting · Internal Use Only Page 24

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

connect to the slide message.

Interest is not the criterion. Relevance to the argument is.

3 The Transition Logic (Slide to Slide)

Each slide must connect logically to the next. There are three types of connections:
Deductive: 'Given A and B, therefore C.' Each slide is a premise in a logical argument.
Inductive: 'A, B, and C are all true. Therefore the general principle D is true.'
Sequential: 'First do A, then B, then C.' Used in implementation slides.
If you can't articulate the logical connection between slide N and slide N+1, you have a
structure problem.

The SCR Framework — Situation, Complication, Resolution

For any communication that needs to set context before delivering the answer, use SCR:

Element Purpose Example

Situation Establishes the stable context — what 'The company has grown revenue at
was true before the problem emerged 18% CAGR for 5 years, building a
strong market position.'

Complication Introduces the tension — what has 'In the last 4 quarters, margin has
changed or what challenge has declined from 14% to 8% despite
emerged continued revenue growth.'

Resolution Delivers the answer — what needs to 'We need to identify and address the
happen to resolve the complication structural cost drivers before they
erode the business model.'

SCR IN A STEERCO OPENING

Situation: 'The SAP S/4HANA implementation has progressed through Explore phase with
94% of workshops completed
and all KDDs resolved. The project team has demonstrated strong engagement and the
technical architecture is stable.'

Complication: 'However, three integration streams are running 3 weeks behind schedule due
to delayed data availability
from legacy systems. If not addressed, this creates a 6-week risk to the planned go-live
date.'

Resolution: 'This SteerCo is asked to approve a revised data migration timeline and
authorize an additional
resource to the integration workstream, enabling an on-time go-live with contained risk.'

HARTS Consulting · Internal Use Only Page 25

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

Written Communication — The Consulting Memo

The consulting memo is a one-page document that communicates a complete, structured
recommendation. Mastering it makes you effective in any written medium — email, report, brief,
or proposal.

• First paragraph: the governing message — what you're recommending and why, in 3-5
sentences
• Second paragraph: the situation and complication — context that makes the
recommendation necessary
• Third paragraph: supporting argument 1 with evidence
• Fourth paragraph: supporting argument 2 with evidence
• Fifth paragraph: supporting argument 3 with evidence
• Final paragraph: implications, risks, and requested actions

The Most Important Communication Rule

Never make the reader work to find your point.

Every sentence should be doing one of two things: stating a claim or proving a claim.
If a sentence does neither — if it's scene-setting, throat-clearing, or 'interesting context' — cut it.
Consulting communication is ruthlessly purposeful. Every word earns its place.

Course Summary — Structured Problem Solving at a Glance

Module 1 Mindset first. Start with the problem. Separate facts, inferences, and
assumptions. Think in structures, not lists. Kill your darlings.

Module 2 Issue trees break complex problems into MECE components. Build them before
you gather any data.

Module 3 Hypothesis-driven thinking directs analysis. Form a specific, testable hypothesis

and gather only the data needed to test it.

Module 4 The 6-stage process: Problem Definition → Structure → Analysis Plan → Data
Gathering → Synthesis → Recommendation.

Module 5 Gather the minimum sufficient data. Use the analysis planning matrix. Primary
research for opinions; secondary for facts. Triangulate.

Module 6 Synthesis is not summarizing. It is transforming findings into insights using the

HARTS Consulting · Internal Use Only Page 26

Structured Problem Solving & Hypothesis-Driven Thinking Consultant Skills Series · Course 1

'So What' test and logical grouping.

Module 7 Recommendations must be specific, grounded, actionable, prioritized, and

honest about risk. Lead with the governing recommendation.

Module 8 Pyramid Principle: lead with the answer. Use SCR for context. Every slide title is
a complete message. Every word earns its place.

End of Course
Structured Problem Solving & Hypothesis-Driven Thinking · Consultant Skills Series

HARTS Consulting · Internal Use Only Page 27