Skip to content

The Edtech App That Stopped Growing

Downloads keep coming. Revenue doesn't. Find the broken stage.

moderate
9 min read
edtechfunnelretention

The Prompt

Your client runs a vernacular-language exam-prep app (government job exams: SSC, banking, railways) with 8 million downloads and ₹90 crore revenue from a ₹999/year subscription. Revenue has been flat for four quarters despite downloads growing 20% per quarter. Diagnose and fix.

Opening exchange

candidate

Downloads up, revenue flat — the leak is between install and renewal. Before structuring: is the ₹999 price unchanged? Any competitor or exam-calendar shifts — fewer notifications released — that would suppress the whole category?

interviewer

Price unchanged. Exam calendar is normal. Competitors exist but their pricing hasn't moved. The problem is internal.

External causes ruled out in one pass — now the funnel is the whole case.

candidate

Then I'll walk the funnel: install → activation (first meaningful study session) → trial-to-paid conversion → renewal. Flat revenue against compounding installs means one of these stages collapsed. I'd like the stage-wise numbers, this year versus last.

Structure & Hypothesis

The funnel with both years' rates, then the cohort tier under activation — 70% @ 9% vs 30% @ 37% reproduces the 24% exactly.

Analysis & Data

interviewer

Activation is your collapsed stage. Dig.

candidate

A rate that halves while volume doubles is usually a mix problem: the marginal installer differs from the historical one. I'd split activation by acquisition channel and by language. Hypothesis: growth marketing widened targeting — cheaper installs from users outside the core exam-aspirant profile, or from languages where content is thin.

Rate collapse + volume growth → always decompose by cohort before blaming the product.

interviewer

Sharp. 70% of new installs now come from a viral short-video campaign in Bhojpuri and Marathi — languages where the app has only 15% of its question-bank translated. Those users activate at 9%. Hindi-channel installs still activate at 37%.

candidate

So growth bought the wrong audience for the current product — or the right audience for a product that doesn't exist yet. Two honest paths: refocus spend on Hindi-belt channels where the product is complete, or treat Bhojpuri/Marathi as the next product investment since the demand signal is real — 70% of installs is not noise. Given flat revenue and a working Hindi engine, I'd do both sequentially: fix the spend now, build Marathi content next (larger exam population than Bhojpuri), launch properly in two quarters.

Recommendation

Recommend

  • Immediately re-weight acquisition spend toward channels/languages where the product is complete; stop buying installs the product can't serve.
  • Build the Marathi question bank to 90% coverage in two quarters — the campaign proved demand; meet it before re-opening the spend tap.
  • Add an activation-by-cohort dashboard as a permanent guardrail: any channel activating <25% gets auto-flagged before budget scales.
  • Address the renewal softness with exam-cycle-aligned renewal offers (renew before prelims at a discount), but treat it as secondary.

Key Takeaway

What this case teaches

When a funnel rate collapses while volume grows, the cause is almost always who's arriving, not what they find. Decompose by cohort first — and recognize that "bad" traffic is sometimes tomorrow's product roadmap announcing itself.