case study frameworks

Circles method can actually help you. I would say practically it gives you a sequence of things to do. But the danger is when you announce it — the interviewer stops listening to your thinking and starts evaluating your memorisation.

Talvinder Singh, from a Pragmatic Leaders interview prep session

Frameworks exist to organise your thinking under pressure. They do not exist to impress the interviewer. The moment you say “I am going to use the CIRCLES framework” out loud, you have told the interviewer two things: you read Cracking the PM Interview, and you are about to follow a recipe instead of thinking.

The best candidates use frameworks silently. Their answer has structure — visible, clear structure — but the structure comes from the problem, not from a memorised acronym. The interviewer sees organised thinking. They do not see a template being filled in.

This page covers the three frameworks worth knowing, when each one helps, and — more importantly — when to abandon all of them.

CIRCLES: the default framework

CIRCLES was designed specifically for product design questions. It stands for:

Letter	Step	What you do
C	Comprehend the situation	Clarify the question. Ask about users, context, constraints.
I	Identify the customer	Pick a specific user segment. State it explicitly.
R	Report customer needs	What does this segment need? What pain are they experiencing?
C	Cut through prioritisation	Generate solutions, then pick one based on impact and feasibility.
L	List solutions	Flesh out the chosen solution with specifics.
E	Evaluate trade-offs	What are the risks? What are you giving up?
S	Summarise	Wrap up with a recommendation tied to a metric.

In practice, the first three steps (Comprehend, Identify, Report) are the most valuable. They force you to do what most candidates skip: understand the problem before solving it. The last four steps are a structured way to present your solution — useful, but less differentiated.

When CIRCLES works: Product design questions (“Design an app for X”, “Improve feature Y”). The framework maps cleanly to these because they follow a user → problem → solution arc.

When CIRCLES fails:

Strategy questions. “Should PhonePe enter insurance?” does not start with a user — it starts with a market. CIRCLES has no step for competitive analysis or business model evaluation.
Metric questions. “Your retention dropped 15%” requires diagnosis, not user identification. CIRCLES will lead you to propose features when the real answer might be “the marketing campaign ended.”
Estimation questions. “How many ATMs are there in Mumbai?” has no user, no customer need, no solution. CIRCLES is irrelevant.

// scene:

Mock interview. A candidate has been asked: 'How would you improve WhatsApp for Indian small businesses?'

Candidate: “I'll use the CIRCLES framework. C — Comprehend. So the question is about improving WhatsApp for small businesses in India...”

Interviewer (thinking): “(They just narrated the question back to me using a framework label. I know this template. Let's see if they can actually think.)”

Compare with a different candidate answering the same question.

Candidate B: “Before I start — when you say small businesses, are we talking about a kirana store owner in Tier 2 using WhatsApp as their only digital tool, or a D2C brand in Bangalore that already uses WhatsApp Business API?”

Interviewer: “Good question. Let's say the kirana store owner.”

Both candidates are doing the same thing — clarifying the user. One announced the framework. The other just did it. The interviewer prefers the second.

// tension:

The framework is a scaffold. Once the building is up, the scaffold should be invisible.

This is the approach I teach because it works for every case type, not just product design. It has no acronym. It does not need one.

Step 1: Clarify (1-2 minutes). Ask questions to narrow the problem. Which users? Which metric? What constraints? What has been tried? The goal is to turn a vague question into a specific one. “Improve Swiggy’s retention” becomes “Why do new users in Tier 1 cities who order in month 1 not order again in month 2?”

Step 2: Structure (1-2 minutes). Decompose the problem into parts. User journey stages, funnel layers, lever categories — pick a decomposition that fits the problem. State it out loud so the interviewer can follow your logic.

Step 3: Analyze (3-5 minutes). Go deep on each part. Quantify where possible. Prioritise: which part of the structure, if fixed, would move the metric most? Reference real behaviour, not hypotheticals.

Step 4: Recommend (2-3 minutes). One specific recommendation, tied to your analysis. Include: what to do, why this over alternatives, how to measure, and what could go wrong.

This approach is covered in depth on the How to Approach Any Case Study page. What this page adds is the choice between frameworks — when the 4-step is better than CIRCLES, and when neither is right.

The 4C framework: for market and strategy cases

When the question is about markets, competition, or business strategy — not product features — use the 4C framework:

C	Focus	Questions to answer
Company	Your position	What are our strengths? What is our business model? Where are we in the market lifecycle?
Customer	Who buys and why	What segments exist? What are their needs? How do they currently solve the problem?
Competition	Who else plays here	Who are the direct and indirect competitors? What is their positioning? Where are they weak?
Context	External environment	Regulations, market trends, technology shifts, macro-economic factors. What is changing?

When 4C works: “Should Razorpay enter lending?” “How would you position a new ed-tech product against Byju’s?” “What should CRED do after credit card bill payments?” These are business questions disguised as product questions.

When 4C fails: Anything requiring a specific product solution. The 4C helps you understand the landscape but does not help you design a feature or diagnose a metric drop.

Choosing the right framework

This is the meta-skill that separates candidates who have memorised frameworks from candidates who know how to think.

Question type	Best approach	Why
”Design a product for X”	CIRCLES (silently) or 4-step	User → problem → solution arc
”Improve product Y”	4-step	Start with metric, decompose, diagnose, recommend
”Should company Z enter market W?“	4C + recommendation	Market analysis before product thinking
”Metric dropped by X%“	SLICE diagnostic	Segment, layer, inspect, correlate, explain
”Estimate the number of X”	Top-down breakdown	Arithmetic, not frameworks
”Tell me about a time you…”	STAR method	Situation, task, action, result

The pattern: match the framework to the question type, not the other way around. A candidate who applies CIRCLES to a metric drop question is telling the interviewer they have one tool and use it for everything.

// thread: #pm-interviews — Post-mock-interview debrief in a Pragmatic Leaders cohort

Ananya I froze on the strategy question. I started with CIRCLES and realised halfway through that I was trying to identify a 'customer' for a market entry question.

Vikram Same thing happened to me last week. The interviewer asked about competitive positioning and I was trying to do user personas.

Rohan (Mentor) This is exactly why we don't teach one framework. Read the question type first. Market question → 4C. Product question → 4-step. Metric question → SLICE. The first 10 seconds decide which lens you use.

Ananya So I need to learn all three?

Rohan (Mentor) Learn the pattern, not the letters. After 10 mocks you won't think about framework names. You'll just structure naturally. 💯 4

When to abandon the framework entirely

Some interview questions break all frameworks. Recognising these and adapting is what gets you a “strong hire” instead of a “hire.”

The vision question. “You are the PM for PhonePe. What is PhonePe’s next big bet?” This is not asking for analysis. It is asking for conviction. Lead with a thesis. “PhonePe has 400 million users who trust it with money but use it three times a month. The next bet is daily commerce — not more financial products.” Then defend it. No framework will generate a thesis for you.

The ethical dilemma. “Your A/B test shows that a slightly misleading CTA increases conversions by 12%. Ship it?” No amount of structuring resolves a values question. State your position, acknowledge the trade-off, explain how you would handle the organisational pressure.

The firefighting case. “Ola’s cancellation rate spiked 40% this week. What do you do?” Time pressure means you skip the structure step. Hypothesise the three most likely causes, propose an immediate investigation plan, define what “resolved” looks like. Speed beats structure here.

In all three cases, the interviewer is testing whether you can think beyond templates. The candidates who reach for a framework on a vision question tell the interviewer they need a recipe. The candidates who lead with judgment tell the interviewer they can handle ambiguity.

// exercise: · 15 min

Framework selection drill

For each question below, decide which approach you would use — and why. Write one sentence for each.

“How would you design a carpooling feature for Google Maps in India?”
“Meesho’s D7 retention dropped from 35% to 22% last month. Diagnose.”
“Should Zomato launch a grocery delivery service?”
“How many Swiggy delivery partners are active in Bangalore on a Friday evening?”
“You are the PM for CRED. The CEO wants to add a social feed to the app. You think it is a bad idea. How do you handle this?”

Check yourself: If you chose the same framework for more than two of these, you are defaulting instead of selecting. Each question demands a different approach.

// interactive:

The framework trap

You are in a PM interview at a fintech company. The interviewer asks: 'India's UPI transaction volume has been growing at 40% year-over-year. If you were the PM for Google Pay, what would you prioritise for the next year?' You have 12 minutes.

The question is ambiguous — it could be a product question, a strategy question, or a market analysis question. How do you begin?

// learn the judgment

You are in a PM interview at CRED. The interviewer says: 'Design a feature that helps CRED users manage their credit card bill payment stress.' You have been practicing CIRCLES for three months. You feel the urge to run through Comprehend Situation → Identify Users → Report User Needs → Cut through prioritization.

The call: Do you follow CIRCLES out loud, adapt it silently, or abandon it for a different structure?

// practice for score