You've memorized the frameworks. You know the acronyms. Now let's see how this actually works with real prompts and example answers.
I'm going to walk through three common case study types with detailed responses—not to give you scripts, but to show you what good thinking looks like in action.
Type 1: Product Improvement
The Prompt: "How would you improve Instagram Reels?"
A Strong Response Structure
Step 1: Clarify and Align (30 seconds)
"Before I dive in, I'd like to understand the focus. Is there a particular user segment or metric you'd like me to prioritize? Also, should I assume we have Instagram's full resources, or are there constraints I should keep in mind?"
Why this works: Shows you don't make assumptions. Gets alignment before you waste time going the wrong direction. Also shows you think about constraints.
Let's say the interviewer says: "Focus on creators, and assume we have normal resources for a feature team."
Step 2: Understand the Product (1-2 minutes)
"Let me start with how I understand Reels. It's short-form video competing primarily with TikTok. The core user segments are creators making content and viewers consuming it. Success metrics would include creator posts (supply), viewer engagement (demand), and time spent watching.
For creators specifically, the journey is: discover the format, create first Reel, get feedback/engagement, create more. The key friction points are usually creation tools, discovery of new creators, and feedback on performance."
Why this works: Demonstrates you understand the product, market, and user segments before proposing solutions.
Step 3: Identify Problems (2-3 minutes)
"For creators, I see a few key problem areas:
First, discoverability. New creators struggle to get views. TikTok's algorithm is famously good at surfacing new creators; Instagram has historically favored established accounts.
Second, creation friction. While Instagram has added editing tools, the gap with dedicated apps like CapCut is significant. Creators often edit elsewhere and upload.
Third, feedback and growth. Creators want to understand what's working. Instagram's analytics exist but aren't as actionable as they could be.
I'd like to focus on the discoverability problem since it has the highest leverage—without views, nothing else matters for new creators."
Why this works: Shows structured thinking, prioritizes rather than listing everything, and explains the reasoning.
Step 4: Propose Solutions (3-4 minutes)
"For improving creator discoverability, I'd consider three approaches:
Option A: New Creator Boost Program
Guarantee first-time Reel posters a minimum number of impressions. This removes the 'posting into the void' anxiety and gives the algorithm enough signal to determine content quality.
Option B: Niche Discovery Features
Add a 'Discover New Creators' tab specifically surfacing creators with under 10K followers. Let viewers who enjoy finding new voices opt into this feed.
Option C: Collaboration Matching
Help new creators collaborate with established ones through a matchmaking feature. This gives new creators access to existing audiences.
I'd prioritize Option A because it directly addresses the core pain point with lowest complexity. It's also how we'd gain signal about creator quality—if boosted Reels still don't get engagement, we learn the content isn't resonating."
Why this works: Multiple options show you can think broadly. Clear prioritization shows you can make decisions. Explains why the chosen option is best.
Step 5: Success Metrics (1 minute)
"I'd measure success by:
- •Primary: Number of new creators who post 3+ Reels in first month (activation)
- •Secondary: 30-day creator retention rate
- •Guardrail: Viewer engagement rate doesn't decrease (ensures we're not flooding feed with low-quality content)
We'd A/B test this with a cohort of new creators over 4-6 weeks."
Why this works: Shows business thinking, includes guardrails (not just optimizing one metric), and mentions testing.
Type 2: New Product Design
The Prompt: "Design a product to help people reduce screen time."
A Strong Response Structure
Step 1: Ask Clarifying Questions
"A few questions to frame my approach:
- •Is this for a specific platform (Apple, Google, standalone app)?
- •Are we targeting a specific demographic?
- •Should I focus on particular types of screen time (social media, gaming, all phone use)?
- •What's our business model assumption?"
Let's say the interviewer says: "Standalone app, focus on young adults and social media, subscription model."
Step 2: Define the User and Problem
"Let me start with who we're building for. Young adults—let's say 18-30—who feel they spend too much time on social media and want to reduce it.
The key insight is: most people already know they're spending too much time. The problem isn't awareness—it's action. They set limits, ignore them, and feel guilty. The tools built into iOS and Android are easy to bypass.
Our users' jobs-to-be-done: 'Help me actually stick to my screen time intentions when my willpower fails.'"
Step 3: Explore Solution Space
"Let me think through approaches:
Approach A: Make it harder to cheat
Lock apps in ways that are genuinely difficult to bypass. Require a friend to unlock, or build in time delays before you can override.
Approach B: Make alternatives more appealing
Instead of just blocking, suggest what to do instead. 'You've hit your Instagram limit—here's a 5-minute meditation instead.'
Approach C: Social accountability
Connect with friends who share your goals. See each other's screen time. Gentle peer pressure and support.
Approach D: Financial stakes
Deposit money that you lose if you exceed limits. Like Beeminder but for screen time.
I think a combination of A and C is most promising. Pure willpower tools fail. Social accountability provides sustained motivation. Making it harder to cheat addresses the moment of weakness."
Step 4: Define the Core Product
"The core product would be:
Commitment contracts with friends: You and a friend both commit to screen time limits. You can see each other's progress. If you want to override, your friend gets a notification.
Escalating friction: First violation is easy to override. Second requires a 5-minute delay. Third notifies your accountability partner. Fourth requires their approval.
Weekly check-ins: The app prompts you and your partner to discuss how the week went. What was hard? What helped?
This is fundamentally a social product that happens to use technology to support behavior change."
Step 5: MVP and Metrics
"For MVP, I'd start with just the accountability partner feature for Instagram (most problematic app for our demographic).
Success metrics:
- •Primary: Weekly screen time reduced vs. baseline
- •Secondary: 30-day retention
- •Leading indicator: Users who complete accountability partner setup
We'd need to watch for users gaming the system or relationships being damaged by the accountability dynamic."
Type 3: Analytical/Metric Diagnosis
The Prompt: "Our e-commerce site's conversion rate dropped 15% last week. How would you investigate?"
A Strong Response
Step 1: Clarify Scope
"A few clarifying questions:
- •Is this 15% relative (from 3% to 2.55%) or absolute?
- •Did traffic volume change significantly?
- •Any known changes last week (releases, pricing, marketing campaigns)?
- •Is this across all devices and geographies?"
Let's say: "15% relative drop, traffic actually increased, no known releases, happening across all segments."
Step 2: Structure the Investigation
"I'd structure my investigation in three phases: data exploration, hypothesis generation, and hypothesis testing.
For data exploration, I'd want to break down the funnel:
- •Did traffic quality change? (traffic source mix, new vs. returning)
- •Where in the funnel is the drop? (home → product → cart → checkout → purchase)
- •Is it across all products or specific categories?
- •Is it consistent across devices?
Let me walk through likely hypotheses for each funnel stage."
Step 3: Walk Through Hypotheses
"If the drop is early in funnel (home → product):
- •Traffic source mix changed (more low-intent traffic from paid)
- •Homepage or navigation changes affected browsing behavior
- •Site performance issues increased bounce rate
If the drop is mid-funnel (product → cart):
- •Price changes or competitor pricing
- •Product page experience issues
- •Out-of-stock items not clearly indicated
- •Review or rating changes
If the drop is late-funnel (cart → purchase):
- •Checkout flow changes
- •Payment issues (new processor, declined cards)
- •Shipping costs or delivery time changes
- •Discount code issues
If traffic increased but conversion dropped:
- •Very likely a traffic quality issue
- •Check: where did the new traffic come from?
- •New marketing campaign bringing less qualified visitors?"
Step 4: Prioritize Investigation
"Given that traffic increased while conversion dropped, I'd start by investigating traffic quality. That's the most likely culprit.
Steps I'd take:
- •Segment conversion by traffic source. Did the new traffic have lower baseline conversion?
- •Compare conversion rates for traffic sources against previous weeks
- •If traffic source conversion rates are stable, the issue is on-site
If traffic quality isn't the issue, I'd move to technical investigation—site speed, checkout errors, payment failures—since those would affect all users regardless of source."
Step 5: What Data Would You Need
"To run this investigation, I'd need:
- •Daily conversion rates by traffic source for past 4 weeks
- •Funnel conversion at each step by day
- •Session recordings or heatmaps for checkout flow
- •Error logs from checkout and payment systems
- •Any scheduled changes from engineering or marketing teams
First priority: get the traffic source breakdown. That will tell us if this is a traffic problem or a site problem."
Final Tips for Case Studies
- •
Think out loud: Silence is your enemy. Share your thinking as you go.
- •
Structure before content: Set up your framework before diving into answers.
- •
Trade depth for breadth: Better to go deep on one strong solution than shallow on five.
- •
Connect to business goals: Always tie back to why this matters for the company.
- •
Acknowledge uncertainty: "I'd want to test this assumption" shows maturity.
- •
Practice with a timer: Most case studies are 15-20 minutes. Know how long your answer is taking.
These examples aren't scripts to memorize—they're patterns to internalize. The structure stays the same; the content changes based on the prompt. Practice the structure until it's automatic, and you'll be able to adapt to any case they throw at you.