Case study · Edtech · Our product

Coachify: how we built an MBA entrance prep platform from scratch and grew it to 7,000+ students

Coachify helps students prepare for CAT and ten other management entrance exams. We designed and built the whole platform, from courses and mock tests to AI doubt solving, payments and the admin team’s tools, and we run its growth through search, content and automation.

The problem
Students prepared across many websites and apps and waited hours for answers to their doubts, while big brands owned the search results and most visitors left without buying.
What we built
One platform for courses, real past papers, AI doubt solving, IIM predictors and payments, plus the admin tools the team runs it on.
The result
+72% clicks from Google and twice the visits in Aug–Oct 2026 compared with a year earlier, and 7,000+ students on the platform.
Product
Coachify (coachifylive.com)
Industry
Edtech, MBA and BBA entrance preparation
Our role
Product strategy, design, engineering, SEO and AEO, growth
Timeline
January 2024 to today, still in active development
Platform
Web app for students, admin panel for the team

At a glance

  • +72%more clicks from Google searchAug–Oct 2026 vs Aug–Oct 2025, Google Search Console
  • 2×daily visits, with 2.5× more usersSame periods, Google Analytics
  • 7,000+studentsAs stated on coachifylive.com
  • 11entrance exams coveredCAT, XAT, SNAP, CMAT, IPMAT and more
Coachify homepage on a laptop: course offers, the IIM eligibility checker and a free counselling form
Coachify homepage on a phone

01 / Background

Background

Every year, a few lakh students in India prepare for CAT and the other management entrance exams. Most of them are in college or working full time, and they study in the gaps: late at night, on weekends, on their phones. The market is crowded. A handful of large coaching brands own most of the search results, the YouTube channels and the ad space.

Coachify wanted to compete with a smaller team and a smaller budget, by being more useful. That meant one platform where a student can learn, practise on real exam papers, clear doubts, check their chances at each IIM and get a mentor, without bouncing between six different websites.

We have built and run Coachify since January 2024. What follows are the problems we found along the way, for students, for the Coachify team and for the platform itself, and what we built to solve each one.

02 / Overview

The problems at a glance

03 / The work

Pain points and how we solved them

01

Preparation was scattered

The problem

A typical aspirant watched lessons on one site, took mocks on another, downloaded past papers as PDFs and asked doubts on WhatsApp. Progress lived nowhere.

What we built

One platform with live classes, recorded courses, mock tests, past papers, free tools, a student community and one-to-one mentoring. Each student gets a dashboard with daily targets and course progress, so they know what to do today.

Outcome

Students can do their whole preparation in one account, and Coachify can see where each student is stuck.

02

Past papers were only PDFs

The problem

Students practised past CAT papers on paper, so they never got used to the real computer-based test, its timer, its on-screen calculator or its sectional time limits.

What we built

Full-length and sectional tests built from previous years’ papers (2019 onwards) for 11 exams: CAT, XAT, SNAP, CMAT, IPMAT Indore, IPMAT Rohtak, JIPMAT, CUET and the IIM Bangalore and IIM Kozhikode undergraduate exams. The tests mirror the exam interface, include the on-screen calculator, and end with a detailed analysis by section and topic. Students can also practise by topic, take a diagnostic test or build a custom test.

Outcome

Students rehearse in the same format as the real exam and see exactly which topics cost them marks.

Coachify’s CAT previous year papers page with slot-wise tests and a quick test maker
CAT past papers as slot-wise tests, with practice by section.
03

Solutions were long and hard to follow

The problem

Explanations for past questions came from many sources, varied in quality and were often too long to revise from.

What we built

An internal tool that uses Google Gemini to rewrite each solution into a short, step-by-step explanation of about 50 to 70 words, with diagrams placed at the right step. The team compares each rewrite side by side with the original before publishing it, and the original records in the database are never changed.

Outcome

Solutions read the same way across every paper and are short enough to revise from.

04

Doubts waited hours

The problem

Students posted doubts in WhatsApp groups and waited for a teacher to reply, often after they had moved on.

What we built

An “ask a doubt” button on every past-paper question. An AI tutor built on Gemini answers in seconds, reads graphs and tables in the question, and sticks to the official solution’s method instead of inventing shortcuts. Students can ask follow-up questions, and generate a fresh practice question on the same concept. Usage is rate limited so AI costs stay predictable.

Outcome

Doubts on past papers get answered while the student is still on the question.

05

Students didn’t know their chances

The problem

Eligibility rules differ across IIMs and other B-schools (category, academic scores, work experience), and students rarely knew which colleges they could target or what a given score meant.

What we built

A set of free tools: an IIM eligibility checker covering 60 business schools, a CAT percentile predictor, a score calculator, a B-school selector and an AI profile evaluator that reviews a student’s background and suggests where to apply. An SOP writer and reviewer helps with applications later on.

Outcome

These tools answer the questions students search for most, and they now bring in more than 10,000 users a month. “CAT score calculator” alone earned 22,700 Google impressions and 766 clicks in the last 16 months.

Coachify’s free CAT percentile predictor: attempts and expected correct answers for VARC, LRDI and QA
The free CAT percentile predictor.
06

Courses were too expensive for many students

The problem

For many students, especially those still in college, the full course fee was the reason they didn’t buy.

What we built

Pool payments. A student pays half the course fee and gets a 15-day invite link to share with a study partner. When the partner pays the other half, the course unlocks for both of them at once.

Outcome

Two students can share the cost of one course without sharing an account.

  1. 1Student Apays half the course fee
  2. 2Invite linkvalid for 15 days, shared with a study partner
  3. 3Student Bpays the other half
  4. 4Unlockedthe course opens for both, each on their own account
How a pool payment works.
07

Students picked the wrong DILR sets

The problem

In the DILR section, choosing which puzzle sets to attempt matters as much as solving them, and students rarely practised that choice.

What we built

CogniCAT, an AI trainer for DILR. Its Set Selector shows a puzzle and gives the student 60 seconds to decide whether to attempt or skip it, then explains the decision. A Puzzle Remixer creates new variations of a puzzle with the same logic, and the tutor gives Socratic hints instead of answers.

Outcome

Students practise the decision that most often costs them marks in DILR, not only the solving.

  1. 1A puzzle setappears, as in the exam
  2. 260 secondsto read it and decide
  3. 3Attempt or skipthe student commits to a choice
  4. 4Explanationwhy that set was worth attempting or skipping
One round of CogniCAT’s Set Selector.
08

Paid content was being shared

The problem

Recorded lessons and notes are easy to download, screen-record or pass around, and one login can be shared by a whole group of friends.

What we built

DRM-protected video through VdoCipher, a limit of two devices per account using device fingerprinting, and a watermark with the student’s identity on paid content that restores itself if someone tries to remove it.

Outcome

Paid content stays with paying students, and a login can’t be used on more than two devices.

09

Visitors left without buying

The problem

Many logged-in students opened a course page, read it and left.

What we built

An automatic follow-up. Two minutes after a logged-in student leaves a course page without buying, they get a WhatsApp message and an email about that course, never more than once a day. The team also runs WhatsApp campaigns from approved templates through WATI, and can schedule them ahead of time.

Outcome

Every interested student gets a timely, relevant nudge without the sales team chasing each one by hand.

10

Nobody knew which channel worked

The problem

Leads came from YouTube, Instagram, search and WhatsApp, but the team couldn’t tell which source brought in paying students.

What we built

Campaign tracking (UTM) on every lead form, saving both the first source that ever brought the student in and the source of the current visit.

Outcome

The team can see which channels and campaigns bring in leads and spend time accordingly.

11

New courses were slow to launch

The problem

Every new batch meant copying lectures into a new course and asking developers for changes.

What we built

A modular course system. A course is a bundle of modules (foundation, advanced, recorded lectures, test series, live classes), and one lecture can belong to many courses without being copied. Access is worked out from completed orders, so a course unlocks the moment payment is confirmed. An admin panel with 130+ screens and role-based permissions lets the content, sales and support teams run all of this themselves.

Outcome

The team launches new courses and offers without waiting on a developer.

12

Big brands owned the search results

The problem

The large coaching brands had years of content and backlinks.

What we built

SEO and AEO in the product itself rather than added afterwards. Structured data on 30+ page templates so Google and AI assistants can read and quote the pages. Separate sitemaps for pages, courses, blogs and past papers, with new pages sent to search engines as soon as they’re published. Search Console clicks and rankings for every blog post inside the admin panel. An automated SEO health check. We also mapped the pages a leading competitor ranked for and built the useful ones Coachify was missing, such as the percentile predictor, the on-screen calculator, formula revision and speed maths.

Outcome

Clicks from Google search grew 72% (Aug–Oct 2026 vs Aug–Oct 2025), the share of searchers who clicked doubled from 0.5% to 1.1%, and Coachify now gets visits from AI assistants like ChatGPT (399 in Aug–Oct 2026).

13

The site slowed down and had security gaps

The problem

By October 2026 the site was slowing to a crawl at busy times. It wasn’t the traffic. 39,000 page views a month were producing 3.1 million server requests, about 80 per page view, and one public API call took 105 seconds and held up everything behind it. The same review found gaps in API access control.

What we built

A full speed and security audit, fixed without touching the payment flow. We also closed about 35 API access gaps and added rate limits on every public, student and admin API.

What we measuredBeforeAfter
Slowest public API call~105 s19 ms
Homepage on a phone20.1 MB10.3 MB
Homepage images13.0 MB2.9 MB
API calls per 3 page views2917

Outcome

The platform stays fast at busy times and still runs on a single small server (1 vCPU, 2 GB RAM), which keeps hosting costs low.

04 / Results

Results

  • +72% clicks from Google searchAug–Oct 2026 vs Aug–Oct 2025, Google Search Console
  • Visits doubled from 13,600 to 27,038, and users grew 2.5× from 5,532 to 13,627Same periods, Google Analytics
  • 30,700 clicks from Google search in the last 16 monthsGoogle Search Console
  • 399 visits from AI assistants like ChatGPTAug–Oct 2026, Google Analytics
  • 10,000+ monthly users of the free toolsCoachify team
  • 7,000+ studentsAs stated on coachifylive.com

05 / Timeline

How it came together

Dates are when each part first shipped, from the project’s commit history.

  1. First version live
  2. Blog and SEO content
  3. Past-paper practice
  4. Online payments
  5. DRM video and WhatsApp automation
  6. AI features on Google Gemini, student community, automatic follow-ups
  7. Test series
  8. CAT percentile predictor
  9. IIM eligibility checker
  10. Pool payments
  11. Live classes
  12. Speed and security audit

06 / Stack

Tech stack

  • Next.js
  • React
  • Node.js
  • MongoDB
  • Google Gemini
  • Cashfree
  • VdoCipher
  • AWS S3
  • WATI (WhatsApp API)
  • Agenda job queue
  • DigitalOcean

07 / For you

What this means for your product

We built Coachify the way we build client products. We start from the real problems of the people using it, ship in small steps, put growth into the product from day one, and keep it fast and secure as it scales, on a sensible budget.

Planning a platform like this?

Let’s talk →

08 / Inside the product

A closer look at Coachify

Screens from coachifylive.com. Tap or click any screen to see it larger.

Homepage
Percentile predictor on a phone
CAT percentile predictor
CAT past papers as tests
Homepage on a phone
Free CAT daily target
CAT formulas revision
Speed maths on a phone
CAT on-screen calculator
Speed maths practice
Free study materials

More work

More products we built