Unsolicited Growth Analysis · MathonGo Ecosystem · April 2026

Why 4.5L Subscribers
Aren't Converting

A diagnosis of the YouTube → paid funnel across getmarks.app, mathongo.com, and quizrr.in — with 3 experiments, a landing page teardown, and one structural change.

Prepared by
Yash — Founder's Office Applicant
Based on
Live site audit, not assumptions
Horizon
14-day execution window
Traffic exists. The conversion path doesn't.

MathonGo's YouTube channel is clearly working — 4.5 lakh subscribers don't lie. The content is strong, the teacher credibility is real (Anup Sir, 14+ years, 1000+ JEE ranks), and the social proof is exceptional (2,017 students at 99+ percentile, 192 at 100 percentile in a subject).

The problem is what happens after someone watches a video. The path from viewer → paid student is broken at multiple points across all three platforms. This analysis documents exactly where, and what to do about it.

Three surfaces. One ecosystem. Different gaps.

All three platforms are owned by Scoremarks Technologies Pvt Ltd. They serve different functions in the funnel — but the handoffs between them are invisible to users.

getmarks.app
Top-of-funnel practice
1M+ downloads, 2M+ questions solved daily. But the website is a dead end — just two download buttons. No pricing, no course info, no conversion path for someone who isn't ready to download.
Dead end LP
mathongo.com
Paid crash courses
Decent structure, real social proof, good course catalogue. But pricing is hidden behind clicks, hero copy is vague ("Math is when done with MathonGo"), teacher credibility is below the fold, and there's no urgency mechanism.
Conversion gaps
quizrr.in
Paid test series
Strongest LP of the three. Has real testimonials with percentiles, impact stats, feature descriptions. But the money-back guarantee is a disappearing top strip, no pricing on homepage, no lead capture, demo test isn't used as a funnel.
Underutilised moat
Three specific things to run right now.

Experiment 01
Show the 15-page analysis on-screen mid-video
Quizrr's biggest differentiator — the 15-page post-test analysis — is mentioned on the website but never shown in action on YouTube. Edit the top 15 videos by watch-time: at the ~45% mark (peak retention, before drop-off), cut to a real student's analysis report on screen. Flag specific weak topics, time-per-question data, percentile estimate. CTA: "Get your own analysis — free demo test → quizrr.in/analysis-demo." Pin in comments with UTM ?src=yt-analysis-drop.
Target CTR
≥ 3% on card
Demo → Paid
≥ 10%
Projected Enrollments
~45 / month
Measure via
UTM + GA4
Experiment 02
MARKS App as the low-friction bridge to Quizrr
YouTube CTAs currently push to paid courses — high friction for a cold viewer. Instead, redirect to MARKS App download (free, zero commitment). Inside the app: Day 3 push notification → "Ready to test under real exam conditions? Try a free Quizrr test." Day 5 → limited offer. This leverages the existing 1M+ download base and creates a warm funnel rather than asking cold viewers to pay immediately.
App Install Rate
Track via UTM
Install → Paid
≥ 2% target
Why it works
Zero-cost entry
Measure via
App events + Quizrr enrollments
Experiment 03
Pinned comment A/B — generic vs. hard numbers + urgency
Split 20 videos into two groups. Group A: generic CTA ("Check out our test series →"). Group B: "2,017 students hit 99+ percentile using this series last year. Batch closing Friday → [link]." Refresh the urgency trigger every 72 hours. Quizrr's social proof numbers are exceptional — they're just sitting unused on YouTube.
Target CTR (B vs A)
3× vs 1%
Impressions
~200K over 14d
Incremental clicks
~4,000 delta
Measure via
UTM per group
Fix the conversion layer before adding more traffic.

With ₹2L and 14 days, improving the landing page is the highest-leverage move. The acquisition engine works — the conversion layer doesn't. Paid ads into these pages would be wasteful. Better content would have a lagged payoff.

Element Current State Fix
Hero copy (MathonGo) "Math is when done with MathonGo" — vague, no outcome "Get 99+ percentile in JEE Math — 2,017 students did it last year"
Pricing visibility Hidden behind course clicks on both sites Show price range on homepage cards — removes friction, pre-qualifies intent
Money-back guarantee Disappearing top strip on Quizrr, invisible on scroll Sticky persistent badge near every CTA — it's a massive trust signal being wasted
Teacher credibility Anup Sir's bio is below the fold on MathonGo Move above fold. "14 years. 1,000+ JEE ranks." is a headline, not a footnote
WhatsApp support Absent on both sites Add WhatsApp chat widget — biggest drop-off reason in edtech is unresolved objections
getmarks.app LP Two download buttons, nothing else — dead end Add a bridge: "Already using MARKS? Upgrade to Quizrr test series →"
The one structural change.

Single Biggest Lever
Make the demo test the primary conversion funnel — not a feature.

Right now the free demo test on Quizrr exists but it's treated as a product preview. A student completes it and lands on a generic "Buy Now" page. The most powerful asset MathonGo has — the 15-page analysis — is being shown as a sample instead of being used as a personalized sales moment.


The fix: After every demo test, generate a real partial analysis — actual weak topics flagged, time-per-question data, estimated percentile vs past Quizrr users. End the report with one specific, contextual CTA: "You're losing marks in Integral Calculus and Modern Physics. Our JEE Main Test Series has 14 targeted mocks for these exact areas. ₹200 off for the next 24 hours."


The moment a student sees their own weak areas called out with specificity, the product sells itself. You've moved from "trust us, it works" to "look, here's your exact problem — we solve this."

15%+ Target demo → paid rate
3–4× Personalized vs generic offer lift
Zero Extra ad spend required

Every insight in this document comes from actually reading the three live sites — not from assumptions or generic growth frameworks. The gaps are real, the fixes are specific, and the experiments are executable in 14 days without additional headcount.