Applied: One Business Problem Structured Three Ways
Before structure, “sales are down” becomes a noisy room of opinions - pricing, ads, discounts, delivery, competitors, inventory, the weather. After structure, the same problem becomes three clean lenses: where money leaks, where customers drop, and where the operating model cannot keep up.
- One good consultant can structure the same problem in multiple valid ways - the best structure depends on the decision needed.
- Use a profit tree when the issue is margin, revenue or unit economics.
- Use a customer journey funnel when the issue is acquisition, conversion, retention or experience.
- Use an operating-model lens when the issue is execution - people, process, technology, incentives or capacity.
- A structure must be MECE: no overlapping buckets, no missing major causes.
- The winning answer does not stop at the framework. It says, “Here is the structure I would use first, and here is why.”
- The final synthesis should turn analysis into a decision: fix, scale, exit, invest, redesign or test further.
Big Picture - The Shift from Listing Ideas to Structuring Decisions
A weak answer lists everything that might matter. A consulting-grade answer chooses the lens that best explains the problem, then uses the other lenses to challenge blind spots. If you want the broader context of why consultants convert ambiguity into decision clarity, revise What Management Consulting Actually Is.
Core Explanation - One Problem, Three Strong Structures
Take one business problem: a retail chain has flat sales, rising costs and falling profit. A candidate may be tempted to throw frameworks at it. A consultant first asks: “What decision are we trying to make?”
The same problem can be structured three ways:
Structure 1 - Profit Tree for Financial Diagnosis
The profit tree is the fastest way to diagnose a problem stated in financial language. It breaks profit into revenue and cost, then keeps decomposing until you find the driver that moved.
For the retail problem, start here:
- Profit = Revenue - Cost
- Revenue = Footfall or traffic x Conversion rate x Average order value
- Cost = Fixed costs + Variable costs
- Variable costs may include product cost, delivery cost, payment cost, returns and store-level labour linked to sales.
Assume a retailer had revenue of 100 units, cost of goods sold of 65, store costs of 25 and marketing of 5. Profit was 5 units, so profit margin was 5 percent. If revenue stays at 100 but cost of goods sold rises to 70, profit becomes 0. The first hypothesis is not “marketing is weak”; it is “gross margin pressure has erased profit.”
When to use it first: use the profit tree if the interviewer says “profitability,” “margin pressure,” “unit economics,” “cost structure,” “cash burn” or “revenue decline.”
Structure 2 - Customer Journey Funnel for Demand Diagnosis
The customer journey funnel is better when the problem is about growth quality. Sales may be flat because fewer people know the brand, fewer visitors convert, fewer buyers repeat, or loyal customers trade down.
For the retail chain, customer-side hypotheses may look like this:
- Awareness issue: fewer new customers are entering the funnel.
- Consideration issue: customers compare but prefer competitors on assortment, price or convenience.
- Purchase issue: store experience, checkout, stock availability or app friction blocks conversion.
- Repeat issue: customers buy once but do not return because value, service or relevance is weak.
- Advocacy issue: the brand is not creating enough word-of-mouth or habit.
When to use it first: use the funnel if the interviewer says “customer acquisition,” “conversion,” “churn,” “retention,” “brand preference,” “NPS,” “digital journey” or “loyalty.”
Structure 3 - Operating Model for Execution Diagnosis
The operating model asks whether the company has the capabilities to deliver the strategy. This lens matters because many strategies are correct on paper and fail in execution.
For the retail problem, operating-model buckets could be:
When to use it first: use this lens if the issue is “scale,” “implementation,” “service consistency,” “digital transformation,” “turnaround execution” or “post-merger integration.”
Definitions You Should Be Able to Say Cleanly
- Business problem: a measurable performance gap that requires a management decision.
- Problem structuring: breaking an ambiguous issue into logical buckets, hypotheses and analyses.
- MECE: mutually exclusive, collectively exhaustive - no overlapping buckets and no missing major causes.
- Hypothesis: a provisional answer tested through data, logic and evidence.
- Synthesis: the final judgment that converts analysis into a recommendation.
Mini Case Study - DMart Ready: One Growth Problem, Three Structures
DMart Ready shows how an Indian retailer can examine online grocery growth through profit, customer and operating-model lenses rather than copying a single market playbook.
Situation: Indian grocery retail has a difficult online economics problem. Customers value low prices, wide availability and convenience, but picking, packing, delivery, returns and substitutions can quickly eat into margins. For a value retailer, the strategic question is not simply “Should we go online?” It is: How do we grow online without breaking the low-cost promise?
The move: DMart Ready is a useful case because the online channel can be analysed as an extension of a value-led retail model rather than a pure speed race. The primary driver is strategic fit - online growth must reinforce the retailer’s cost discipline and value proposition. Supporting drivers include controlled assortment, fulfilment discipline, geographic focus, inventory visibility and customer pickup or delivery choices.
The lesson: A shallow answer says, “DMart should invest more in e-commerce.” A complete consulting answer structures the same decision three ways:
Outcome or lesson: The best structure protects the core business model while exploring growth. The primary driver is not “online demand exists”; it is whether the channel can grow with unit economics and operating discipline. That is exactly the kind of multi-lens thinking interviewers reward.
How AI Changes Business Problem Structuring
AI does not remove the need for structure. It raises the bar. If every candidate can generate a generic framework, the differentiator becomes choosing the right structure, asking better questions and spotting bad assumptions.
- AI accelerates first-draft issue trees. Tools like ChatGPT or Claude can generate five possible structures for “retail profitability decline” in seconds. Your job is to reject overlap, remove weak buckets and choose the structure that matches the decision.
- AI improves evidence scanning. Perplexity can help scan public filings, news and competitor pages to identify possible hypotheses. But you must separate sourced facts from plausible-sounding filler.
- AI changes consulting economics. Analysts can do faster desk research, slide drafting and benchmark synthesis, which means interviews increasingly test judgment, not just effort. For the consulting-firm angle, revise How AI Is Reshaping Consulting Work and Firm Economics.
Load this lesson, a company annual report and two competitor notes into NotebookLM. Ask: “Give me three different structures for the same business problem, list the assumptions behind each, and suggest which one I should open with in an interview.” Then edit the answer until it is MECE and decision-led.
Interview Relevance
“A retail chain has flat revenue and falling profitability. Structure the problem. Now show me another way to structure the same problem.”
When asked for another structure, do not repeat the same tree with different labels. Change the lens. Move from “profit” to “customer journey” or from “customer” to “operating model.”
Common Mistake
The biggest mistake is framework spraying - naming profitability, 4Ps, SWOT and customer journey without explaining which one fits the decision. It costs candidates because it signals memorisation, not judgment. The one-line fix: “I will start with this structure because the client’s decision is X; I will use the second lens only to check Y.”