growth & acquisition cases

Growth case studies are where most PMs reveal they have never actually managed a funnel. They say “improve acquisition” and propose Instagram ads. They say “improve retention” and suggest push notifications. Neither answer demonstrates that they understand how users move from stranger to customer to advocate — or where that movement breaks.

I have reviewed thousands of growth cases across Pragmatic Leaders cohorts. The ones that fail share a pattern: they treat growth as a marketing problem instead of a product problem. The ones that succeed treat growth as a series of conversion gates, each with its own friction, its own metric, and its own lever.

Here are four fully worked cases. Each dissects a real growth challenge — acquisition, activation, retention, or referral — with the funnel logic, the unit economics, and the India-specific context that makes it real.

The growth funnel in 90 seconds

Before the cases, internalize this skeleton. Every growth case maps to one or more stages:

This is the AARRR framework, proposed by Dave McClure of 500 Startups. It is not complicated. But most candidates cannot tell you which stage their proposed solution actually targets — which means they cannot tell you whether it will work.

The critical insight: these stages are sequential, not parallel. Pouring money into acquisition when your activation rate is 15% means you are paying to fill a leaking bucket. Fix the leak first. Then pour.

Case 1: PhonePe’s activation problem in tier-2 cities

PhonePe has 400 million registered users. But “registered” does not mean “active.” In tier-2 and tier-3 cities, a massive cohort downloads the app because a shopkeeper tells them “UPI se karo” — they install, link their bank account, make one payment, and never open the app again.

The question is not “how do we get more users?” PhonePe is already pre-installed on many Android devices. The question is: why do users who install never reach the aha moment?

Clarifying the problem

This is an activation case, not an acquisition case. The funnel looks like:

The biggest drop is between first and second transaction. That is the activation gap.

The solution: Contextual onboarding by city tier

Instead of the generic “Send money / Pay bills / Recharge” home screen, surface the use case most relevant to the user’s context:

• Tier-2 user, first week: Show electricity bill due date (pulled from BBPS data), prepaid recharge reminder, and local DTH renewal — not UPI send money, which they already did.
• Nudge timing: Send a notification 2 days before the user’s typical recharge cycle (detectable from the first transaction amount pattern).
• Social proof: “42 lakh people in Lucknow paid their electricity bill on PhonePe this month” — localized, specific, credible.

Metric

Activation rate — percentage of new tier-2 users who complete 3+ transactions within 14 days. Current: ~12%. Target: 20%.

Secondary: time-to-second-transaction. If the nudge works, this drops from 11 days to under 5.

Trade-offs

The key insight: The product is the same. The activation gap is not a product quality problem — it is a use-case discovery problem. Tier-1 users stumble into recurring use cases because their lives are already digital. Tier-2 users need the app to show them the job it can do.

Case 2: Meesho’s CAC crisis — when acquisition gets expensive

Meesho grew by making every small-town entrepreneur a reseller. Zero-commission, social-commerce model. Resellers shared product links on WhatsApp, took orders from their network, and Meesho handled logistics.

Then Meesho shifted to a direct-to-consumer model. Suddenly, they needed paid acquisition — and the economics changed overnight.

The numbers that matter

The D2C model has better unit margins per order — but the acquisition cost is 12x higher and retention is half. At 22% D30 retention, most users never place order 5. CAC is never recovered.

Structuring the problem

This is not “how do we reduce CAC” — that is one lever. The real question is: how do we make the unit economics work? There are three paths:

1. Reduce CAC — find cheaper acquisition channels or improve conversion on existing ones.
2. Increase LTV — get users to order more frequently or at higher AOV.
3. Hybrid model — use the reseller network for acquisition (low CAC) and the D2C app for retention and monetization.

Path 3 is the interesting one. Most candidates pick path 1 or 2 in isolation. The insight is that Meesho already has a low-CAC acquisition engine — the reseller network. The mistake was abandoning it, not evolving it.

The solution

• Resellers become acquisition partners, not order processors. Pay resellers Rs 30 per new D2C user they bring (vs Rs 180 on Meta). The reseller shares a referral link instead of taking the order themselves.
• Reseller incentive shifts from margin-per-order to customer-acquisition-bonus + trailing commission. Reseller earns Rs 30 upfront + 2% of the referred user’s orders for 90 days.
• The D2C app handles fulfillment, returns, and retention. Better experience than WhatsApp-mediated ordering. Higher retention because the user has the app installed.

Metric

Blended CAC — weighted average across paid channels and reseller referrals. Target: below Rs 60 (from Rs 180).

Secondary: LTV:CAC ratio at 6 months. Target: 2.5:1 (from 0.7:1).

Trade-offs

The key insight: The cheapest acquisition channel is one where someone who already trusts your product tells someone who trusts them. Paid ads are what you use when you have no trust network. Meesho built a trust network and then stopped using it.

Case 3: CRED’s retention paradox — what do you do after the bill is paid?

CRED has a beautiful app, a premium user base, and a core use case that happens once a month: pay your credit card bill. The product delivers real value — rewards for on-time payment, a clean interface, a credit score tracker. But the problem is frequency.

A user who pays their credit card bill on the 5th of every month opens CRED once. Maybe twice if they check their score. The other 28 days, the app does not exist.

Why this matters

Retention in consumer apps is driven by frequency. Instagram retains because you open it 8 times a day. Swiggy retains because you order 3-4 times a week. CRED retains… once a month. And monthly-active is not enough to justify the CAC CRED pays to acquire premium users.

CRED’s D30 looks healthy — but it is an artifact of the billing cycle, not engagement. The user comes back because the bill is due, not because they want to use CRED. This is clock-driven retention, not value-driven retention. And clock-driven retention does not compound.

Structuring the retention challenge

The fundamental tension: CRED needs to create visit reasons between bill payment dates. The options:

1. Commerce (CRED Store, CRED Cash) — give users reasons to transact between bill payments.
2. Content (CRED Curious, editorial) — give users reasons to consume between bill payments.
3. Gamification (CRED Jackpot, spin-the-wheel) — give users reasons to open between bill payments.
4. Financial hub — expand from credit card bills to rent, insurance, investments — increase the number of monthly financial actions.

The solution: Financial actions, not entertainment

Instead of gamification, expand the number of financial jobs CRED handles:

• Rent payment — CRED already has this, but adoption is low because landlords resist digital payment. The lever: make rent payment via credit card attractive (earn rewards) and make landlord onboarding frictionless (instant bank transfer to landlord, regardless of their tech literacy).
• Insurance renewal reminders — CRED has policy data via account aggregation. Surface renewal dates, compare quotes, let users renew from within CRED. Transforms a once-a-year event into a CRED interaction.
• Subscription management — Show all recurring charges across credit cards. Let users cancel, pause, or switch plans from a single dashboard. This is a weekly use case — “what am I paying for that I do not use?”

Each of these is a real financial job, not a game. Each creates a visit reason tied to genuine value. And each reinforces CRED’s positioning as “the app that manages your financial life” rather than “the app that gives you coins for paying your bill.”

Metric

Monthly active days — number of distinct days per month a user opens CRED. Current: ~2. Target: 6.

Not sessions. Not time spent. Days. Because a user who opens CRED on 6 different days in a month is using it as a financial tool. A user who opens it twice but spends 20 minutes spinning a wheel is not.

Trade-offs

The key insight: CRED’s retention problem is not solvable by making people open the app more. It is solvable by giving people more real reasons to open the app. Gamification adds frequency without value. Financial actions add both. A PM who conflates the two does not understand retention.

Case 4: Zepto’s referral flywheel in a price-sensitive market

Zepto delivers groceries in 10 minutes. Their CAC on paid channels is Rs 350-400. In a category where average order value is Rs 250 and gross margin is under 15%, that CAC is brutal. They need an organic acquisition channel.

Referrals are the obvious answer. But referral programs in India have a specific failure mode: people game them.

The gaming problem

Every Indian consumer app that has launched a referral program has faced this:

• Users create fake accounts using secondary SIM cards to claim referral bonuses.
• Referral fraudsters build WhatsApp groups dedicated to “earn from referral” schemes — the referred user has zero intent to actually use the product.
• The referral bonus gets claimed, the referred user never orders, and the CAC for that “referral” is actually higher than paid ads (because you also paid the referrer).

Structuring the referral design

The question is not “should we have a referral program?” Every competitor does. The question is: how do we design a referral program where the incentive aligns with genuine usage, not just sign-up?

Three design choices matter:

1. When the reward triggers — on sign-up, on first order, or on third order?
2. What the reward is — cash, credits, or product-specific value?
3. Who gets rewarded — referrer only, referred only, or both?

Why this works

• Delayed trigger kills fraud. If the reward requires the referred user to place 3 orders, fake accounts are worthless — nobody is ordering groceries three times on a fake account.
• Free delivery is product-native. Rs 100 cashback attracts deal-seekers. Free delivery attracts people who actually want to order groceries but hate the delivery fee. These are different users.
• Trailing reward aligns incentives. The referrer only gets their bonus if the referred user sticks around for 14 days. This means referrers will refer people who actually need grocery delivery — not random contacts who will never order.

Metric

Referred user D14 retention rate — not just “how many signed up via referral” but “how many referred users are still ordering after 2 weeks.” Target: 35% (vs 12% for typical referral-acquired users in the industry).

Secondary: viral coefficient — average number of activated users each referrer brings. Target: 0.3 (each referrer brings 0.3 active users, meaning every 3 referrers produce 1 new retained user).

Trade-offs

The key insight: In India, referral programs that reward sign-ups attract fraudsters. Programs that reward usage attract customers. The design of the incentive determines the quality of the user it attracts. Most PMs design for volume. Growth PMs design for quality.

The growth case checklist

After thousands of reviews, here is what separates a passing growth case from a strong one:

1. Did you identify the funnel stage? If your answer is about acquisition but the real problem is activation, you solved the wrong thing. State which stage you are targeting and why that stage has the most impact.

2. Did you talk about unit economics? Growth without economics is a fairy tale. Know the CAC. Know the LTV. Know how many orders or months it takes to recover the acquisition cost. If you cannot do this math on a whiteboard, you are not ready for a growth role.

3. Did you distinguish between quantity and quality? More sign-ups is not growth if D7 retention is 8%. More sessions is not engagement if session depth is 40 seconds. The metric must measure the thing you actually want, not a vanity proxy.

4. Did you account for the India context? CAC in India is rising because Meta and Google duopoly controls the ad market. UPI makes payment frictionless but also makes switching frictionless. Price sensitivity in tier-2 is 3-4x higher than tier-1. Referral fraud is endemic. If your growth plan could have been written for a Bay Area startup, it is missing the context.

5. Did you name the second-order effect? Every growth lever has a downstream consequence. Aggressive discounting acquires price-sensitive users who churn when discounts stop. Referral bonuses attract bonus-seekers. Gamification inflates sessions but hollows out engagement quality. Name the second-order effect, or the interviewer will name it for you.

Where to go next

• Master the general framework: How to Approach Any PM Case Study
• Practice product improvement cases: Product Improvement Cases
• Sharpen your metrics thinking: Metrics & Analytics Cases
• Try India-specific challenges: Indian Market Cases