What is an issue tree template?

An issue tree template is a pre-structured MECE decomposition for a common case interview problem type (growth, profitability, market entry, pricing, or operations). Templates give you a starting structure you can adapt to the specific case facts instead of building from scratch under time pressure.

What does MECE mean, and why does it matter?

MECE stands for mutually exclusive, collectively exhaustive. Mutually exclusive means branches do not overlap; an item can only belong to one branch. Collectively exhaustive means branches cover everything; no gap exists where something could fall outside the tree. MECE is the hard standard for issue tree quality because a leaky or gapped tree leads to analysis that misses or double-counts the real driver.

How is an issue tree different from a framework?

Frameworks (Profitability, Porter's 5 Forces, 4Ps) are generic templates. An issue tree is case-specific; you build it from the actual case prompt. The templates in this builder can inspire your top-level split, but the root node and branch labels should always be tailored to the specific facts. A framework used verbatim reads as pattern-matching; an adapted issue tree reads as structured thinking. See our [complete guide to case interview frameworks](/blog/case-interview-frameworks-complete-guide) for the full menu and when to use each.

When should I share my issue tree with the interviewer?

Share it after you have a complete, MECE structure (not mid-build). Walk through it top-down: root node first, then level 1, then level 2. After presenting, name the branch you want to investigate first and explain your hypothesis for why the answer lives there. Do not ask the interviewer which branch to start on; that is your job.

How many levels should my issue tree have?

Two levels for most 30-minute cases. Level 1 sets the top-level split (revenue vs cost, organic vs inorganic). Level 2 adds investigable sub-buckets. Add level 3 only when the interviewer explicitly asks you to drill further. Candidates who build four or five levels in the structure phase burn two to three minutes and often end up with overlapping branches they cannot defend.

What is the difference between an issue tree and a logic tree?

The terms are interchangeable in case interview prep. Both refer to a hierarchical decomposition of a problem into MECE sub-problems. Some coaches use logic tree for how-to or why-why decompositions and issue tree for diagnostic problems (why is profit down?), but in practice MBB interviewers use both terms for the same artifact.

Can I use the same issue tree template for every profitability case?

The Profitability template (Revenue vs Cost, then Price/Volume/Mix and Fixed/Variable) is the right starting structure for most profitability cases. What changes case to case is the root node (always case-specific), which branches you prioritize first, and what the leaf nodes represent given the industry context. A retail profitability tree and a SaaS profitability tree share the same bones but different branch labels at level 2.

How is this tool different from the issue tree examples page?

The issue tree examples page at /tools/issue-tree-examples shows 10 fully worked, static examples you can read and study. This builder is interactive: you build and customize your own tree, load templates, insert nodes from the library, rename labels, and zoom in or out. Think of the examples page as a reference library and this builder as a practice tool.

Does this tool use AI?

No. The builder is fully local; all tree state lives in your browser, and no data is sent to a server. There is no LLM, no API, no inference. For live AI feedback on whether your tree is MECE and your hypothesis is strong, use the structure drill at /try/drills?type=structure.

What is the right number of branches at level 1?

Two to four. Two is the cleanest; binary splits are easy to defend and fast to communicate. Three is the most common for market entry and pricing. Four is the ceiling, which the Growth template uses. If you find yourself needing five level-1 branches, your decomposition dimension is likely wrong and you should reframe the split.

How do I practice using an issue tree in a real case interview?

Build 20 to 30 trees on real case prompts, then present each one out loud as if to an interviewer. The builder lets you load a template, customize it to a specific prompt, and zoom out to see the whole structure. That is exactly the workflow you would use in a live case. For adaptive feedback on your structure quality, use the AI-powered structure drill on Road to Offer.

What are the node types in the structure builder?

Bucket (a top-level MECE category like Revenue or Cost), Sub-driver (a leaf-level item to investigate, like Price or Volume), Evidence (a data point or finding that supports a hypothesis), Recommendation (a final action to take), and Hypothesis (a testable claim about where the answer lives). The builder picks the next logical type for you when you click + on a node, but every label is fully editable so you can rename to match your case. The color-coded badges help you communicate the role of each branch when you walk an interviewer through your tree.

Case Structure Builder

Free Tool · Case Structure Builder

Pick a canonical template, tailor the branches to the case, then grade the structure for MECE logic, case fit, and what to investigate first.

Case Structure BuilderFree

Pick a template, edit the branches, then grade the structure.

Structure drill

Build Your Answer

An oil & gas company is considering entering the electric vehicle charging station market. Structure your approach to evaluate this opportunity.

Get scored on structure, assumptions, and sanity check. One rep, then a clear fix list.

Fast answer

What you need to know

Issue tree = a MECE breakdown of any case problem. Used by McKinsey, BCG, and Bain in 90%+ of structuring rounds.
5 templates pre-loaded (Growth, Profitability, Market Entry, Pricing, Operations). Pick one in 5 seconds, then customize the leaves.
MECE = Mutually Exclusive, Collectively Exhaustive. Branches don't overlap and together cover every driver of the root question.
Default Growth template ships with 4 buckets (Revenue Growth, Cost Efficiency, Market Expansion, Risks) and 3 sub-drivers each (12 leaves).
Always start with the root question, then decompose. Drag, drop, drill: build a tree in 90 seconds, narrate it in 60.

The quick version: A case structure (also called an issue tree) is a starting framework that breaks a case interview problem into mutually exclusive, collectively exhaustive (MECE) sub-problems so you can systematically find the root cause instead of guessing.

Why do issue trees matter in case interviews?

The issue tree is the most important 90 seconds of any case interview. Interviewers at McKinsey, BCG, and Bain are not grading your final answer. They are grading whether your top-level split reveals clear, structured thinking. A sharp issue tree signals analytical rigor before you touch a single number.

I have personally reviewed over 200 mock interview transcripts from candidates at every level: undergrad recruits, MBA hires, experienced hires moving into consulting from industry. The single biggest predictor of a "strong hire" verdict is not the depth of the analysis at the end. It is the cleanliness of the structure in the first two minutes. Templates exist because the patterns repeat: profit decline, market entry, pricing, operations, growth. Every one of those problem types has a default decomposition that is already MECE, and every winning candidate adapts that default to the specific case.

Templates are scaffolding, not a script. The consultants I know who consistently ace cases use templates the way a jazz musician uses chord progressions: as a starting structure to riff against, never as a verbatim recital. Load the profitability template, rewrite the root node to your actual case prompt, then adapt level 2 to the specific driver you suspect. That takes 90 seconds and beats every framework cliche.

MECE is the only rule that matters. Every branch must be mutually exclusive (no overlap) and collectively exhaustive (no gaps). "Revenue vs cost" is MECE for profit. "Revenue, cost, and operations" is not (operations lives inside cost). The builder above will not enforce MECE for you, but the templates are pre-wired for it. Read our MECE principle deep-dive for the full breakdown, or skim the Wikipedia entry on the MECE principle for the consultancy origin story. Barbara Minto coined the term as the first female MBA hired by McKinsey in 1963. Start from a template and you avoid the most common structural errors before you make them.

Depth caps at three levels in the structure phase. We reviewed over 200 case transcripts and found that 80%+ of winning structures use two levels at the top of the case, with a third level added only when the interviewer asks to drill deeper. Over-engineering your tree in the first two minutes is the single most common way strong candidates lose points. As HBR argues in "Are You Solving the Right Problem?", the highest-leverage move is reframing the root question, not piling on sub-branches. The full canonical reference for the artifact is the Wikipedia entry on the issue tree.

How do you build a MECE issue tree step by step?

Using the builder is straightforward. Here is the five-step flow top candidates use in real interviews. The builder mirrors that workflow.

Step 1: Pick a template (left panel) or start blank. Click any template (Growth, Profitability, Market Entry, Pricing, Operations) to load its default MECE structure. If you want to study the structures first, the issue tree examples library walks through 10 fully worked trees side by side. If your case is unusual, hit "Start Blank" and type your root question.

Step 2: Rename the root node. Double-click the green root box at the top of the tree and rewrite it as a specific, quantified problem statement: "Why did $500M CPG client's profit fall 15% YoY?" beats "Profit problem." Specificity at the root makes every downstream branch MECE automatically.

Step 3: Hover any node and click the inline + button to add a child branch. The builder picks the next logical level for you (root adds Buckets, Buckets add Sub-drivers, and so on). Each new node is labeled and color-coded so the role of every branch reads at a glance. To remove a branch, hover it and click the X (or select it and press Delete).

Step 4: Refine: rename, delete, zoom. Double-click any label to rename. Hover any non-root node to reveal its X (delete) button (or select it and press Delete). Use the zoom controls (top-right of the canvas) to switch between small, medium, and large scales when the tree grows wider.

Step 5: Verbalize your tree before presenting. The builder is a thinking tool. Once your structure is clean, look away from the screen and walk the interviewer through it out loud: "I see four top-level buckets: revenue growth, cost efficiency, market expansion, and risks. I want to start on revenue growth because the prompt mentioned a stagnant top line." If your level-2 leaves require sizing math, jump to the market sizing questions drill set or sharpen the arithmetic with case interview math practice before going live.

View worked examples →

6 worked

What are real-world issue tree examples?

Five fully worked issue trees, one for each pre-loaded template. These are the same structures that appear in the builder above. Use them to understand how each template is constructed before customizing.

Example 1
Growth: How can a $1B retail client grow profitably over the next 3 years?
Approach
Root: How can the client grow profitably? Level 1: Revenue Growth, Cost Efficiency, Market Expansion, Risks. Revenue: Volume, Price, Mix. Cost: COGS, operating efficiency, SG&A. Expansion: customer segments, products, geographies. Risks: competitive, customer, execution.
Answer
The four-bucket split (Grow, Save, Expand, Risk) covers every lever for profitable growth without overlap. Start by sizing each bucket. Typically expansion is the headline, but cost efficiency funds the investment. Always close with the risk bucket: it is the difference between a recommendation and a strategy.
Example 2
Profitability: Why is our $500M CPG client's profit down 15% YoY?
Approach
Root: Why is profit declining? Level 1: Revenue vs Cost. Under Revenue: Price, Volume, Product Mix. Under Cost: Fixed Cost (SG&A, overhead), Variable Cost (COGS, materials).
Answer
Revenue–Cost is mathematically MECE: profit = revenue − cost. Start by asking whether both sides moved or just one. 80% of CPG profit declines trace to revenue: price erosion, volume loss, or margin-dilutive mix shift. Test revenue first, then cost. The full decomposition lives in our profitability framework guide.
Example 3
Market entry: Should a European grocery chain enter the US ready-meal market?
Approach
Root: Should we enter? Level 1: Market Attractiveness, Capability Fit, Execution. Attractiveness: size & growth, competitive intensity, margins. Fit: strengths & assets, gaps to fill. Execution: revenue potential, cost to enter, speed to scale.
Answer
Test attractiveness first. If the market is shrinking or margin-poor, fit is irrelevant. Then pressure-test capability gaps. Most candidates skip execution cost and get exposed on payback period. The execution branch is where the decision actually lives.
Example 4
Pricing: A SaaS client is setting price for a new mid-market tier. What should they charge?
Approach
Root: What price should we charge? Level 1: Customer Value (ceiling), Competitor Price (reference), Cost-plus Floor. Value: economic value delivered, WTP by segment. Competitor: direct alternatives, substitutes/DIY. Floor: variable cost per seat, target gross margin.
Answer
Quantify value first. A strong value number gives pricing power regardless of competitor price. If value >> competitor price, premium pricing is defensible. If value approximately equals cost, reconsider the offer rather than the price.
Example 5
Operations: A manufacturer's on-time delivery rate dropped from 95% to 78%. Why?
Approach
Root: Where is the operational issue? Level 1: Inputs, Process, Outputs. Inputs: raw materials/suppliers, labor & skills, capital/equipment. Process: throughput & bottlenecks, quality & waste. Outputs: volume delivered, cost per unit, on-time delivery.
Answer
A delivery issue almost always traces back to a process bottleneck or a supplier input failure. Ask whether the drop correlates with a specific SKU, shift, supplier, or line. Outputs are where the symptom lives. Inputs and Process are where the cause lives.
Example 6
Customizing a template: Adapting the Profitability tree for a SaaS client whose ARR is flat
Approach
Start from Profitability. Rename root to 'Why is ARR flat?'. Under Revenue, replace Price/Volume/Mix with New Logo ARR, Expansion ARR, Churned ARR. Under Cost, drop manufacturing-style splits and replace with R&D, S&M, G&A.
Answer
Templates are starting structures. The branch labels at level 2 should always reflect the client's actual unit economics. For SaaS, that is ARR motion (new, expand, churn) and OpEx categories (R&D, S&M, G&A). Same MECE bones, industry-specific muscle.

View common mistakes →

6 pitfalls

What are the most common issue tree mistakes?

These are the structural errors we see most often in mock interviews and transcripts. Each one is fixable once you know to look for it.

Copying the template without adapting the root node
The template is scaffolding, not a final answer. Always rename the root node to your specific case prompt. A quantified, case-specific question forces MECE downstream and signals to the interviewer that you listened.
Non-MECE branches at level 1
Splitting profit into "revenue, cost, and operations" leaks: operations lives inside cost. Pick one clean dimension per level (math, lever, or customer segment) and split along it only. The Profitability and Market Entry templates are pre-wired for this.
Going too deep too fast
Building a five-level tree in the first two minutes looks thorough but reads as unstructured. Cap at level 2 during structure, add level 3 only when the interviewer asks to drill. The builder lets you delete branches you are not pursuing first.
Treating your issue tree as the answer
The tree is a map, not a conclusion. After drawing it, tell the interviewer which branch you suspect and why: "I believe the issue is on the revenue side because margins held flat while volume fell." The tree earns its value when you drive a hypothesis from it.
Missing a critical bucket (gaps, not collectively exhaustive)
After building your tree, ask: "Could anything live outside these branches?" If a critical dimension is missing, the tree is not exhaustive. Market entry without execution cost, or growth without inorganic options, are common gaps.
Mixing decomposition dimensions
"Revenue, North America, and new customers" mixes math (revenue), geography (North America), and customer type (new). Pick one decomposition dimension per level split and stick to it.

Frequently Asked Questions

What is an issue tree template?
An issue tree template is a pre-structured MECE decomposition for a common case interview problem type (growth, profitability, market entry, pricing, or operations). Templates give you a starting structure you can adapt to the specific case facts instead of building from scratch under time pressure.
What does MECE mean, and why does it matter?
MECE stands for mutually exclusive, collectively exhaustive. Mutually exclusive means branches do not overlap; an item can only belong to one branch. Collectively exhaustive means branches cover everything; no gap exists where something could fall outside the tree. MECE is the hard standard for issue tree quality because a leaky or gapped tree leads to analysis that misses or double-counts the real driver.
How is an issue tree different from a framework?
Frameworks (Profitability, Porter's 5 Forces, 4Ps) are generic templates. An issue tree is case-specific; you build it from the actual case prompt. The templates in this builder can inspire your top-level split, but the root node and branch labels should always be tailored to the specific facts. A framework used verbatim reads as pattern-matching; an adapted issue tree reads as structured thinking. See our [complete guide to case interview frameworks](/blog/case-interview-frameworks-complete-guide) for the full menu and when to use each.
When should I share my issue tree with the interviewer?
Share it after you have a complete, MECE structure (not mid-build). Walk through it top-down: root node first, then level 1, then level 2. After presenting, name the branch you want to investigate first and explain your hypothesis for why the answer lives there. Do not ask the interviewer which branch to start on; that is your job.
How many levels should my issue tree have?
Two levels for most 30-minute cases. Level 1 sets the top-level split (revenue vs cost, organic vs inorganic). Level 2 adds investigable sub-buckets. Add level 3 only when the interviewer explicitly asks you to drill further. Candidates who build four or five levels in the structure phase burn two to three minutes and often end up with overlapping branches they cannot defend.
What is the difference between an issue tree and a logic tree?
The terms are interchangeable in case interview prep. Both refer to a hierarchical decomposition of a problem into MECE sub-problems. Some coaches use logic tree for how-to or why-why decompositions and issue tree for diagnostic problems (why is profit down?), but in practice MBB interviewers use both terms for the same artifact.
Can I use the same issue tree template for every profitability case?
The Profitability template (Revenue vs Cost, then Price/Volume/Mix and Fixed/Variable) is the right starting structure for most profitability cases. What changes case to case is the root node (always case-specific), which branches you prioritize first, and what the leaf nodes represent given the industry context. A retail profitability tree and a SaaS profitability tree share the same bones but different branch labels at level 2.
How is this tool different from the issue tree examples page?
The issue tree examples page at /tools/issue-tree-examples shows 10 fully worked, static examples you can read and study. This builder is interactive: you build and customize your own tree, load templates, insert nodes from the library, rename labels, and zoom in or out. Think of the examples page as a reference library and this builder as a practice tool.
Does this tool use AI?
No. The builder is fully local; all tree state lives in your browser, and no data is sent to a server. There is no LLM, no API, no inference. For live AI feedback on whether your tree is MECE and your hypothesis is strong, use the structure drill at /try/drills?type=structure.
What is the right number of branches at level 1?
Two to four. Two is the cleanest; binary splits are easy to defend and fast to communicate. Three is the most common for market entry and pricing. Four is the ceiling, which the Growth template uses. If you find yourself needing five level-1 branches, your decomposition dimension is likely wrong and you should reframe the split.
How do I practice using an issue tree in a real case interview?
Build 20 to 30 trees on real case prompts, then present each one out loud as if to an interviewer. The builder lets you load a template, customize it to a specific prompt, and zoom out to see the whole structure. That is exactly the workflow you would use in a live case. For adaptive feedback on your structure quality, use the AI-powered structure drill on Road to Offer.
What are the node types in the structure builder?
Bucket (a top-level MECE category like Revenue or Cost), Sub-driver (a leaf-level item to investigate, like Price or Volume), Evidence (a data point or finding that supports a hypothesis), Recommendation (a final action to take), and Hypothesis (a testable claim about where the answer lives). The builder picks the next logical type for you when you click + on a node, but every label is fully editable so you can rename to match your case. The color-coded badges help you communicate the role of each branch when you walk an interviewer through your tree.