Rebuilding Retention During Vedantu's COVID Surge
Ran structured A/B tests on manual customer success to decide how to scale support during a demand spike.
- Customer Retention
- A/B Testing
- Ed-Tech
- Customer Success
- Product Onboarding

Ujjawal Rahangdale
Associate Director at Singularium Technologies





From their time as

Senior Program Manager - COO's Office
Vedantu • 2020 - 2021
Overview
Ujjawal ran a controlled test comparing manual customer success agents against a ticketing system during a COVID-era demand surge at Vedantu, a decision that reshaped how the company approached support at scale. The test informed a broader retention rebuild, and 30-day retention moved from 33% to 45% during that period.
The Story
Ujjawal joined Vedantu's COO office at a moment when a sudden shift to online learning during COVID had pushed demand far beyond what the existing support setup was built for. Free and paid students were seeing the same landing page, and the team had no clear read on what was actually driving drop-off among the platform's core JEE prep cohort.
He worked with the product team to redesign the landing page and user journey through A/B testing, while separately designing an experiment on the customer success side. He organized JEE cohort students into WhatsApp batches, placing a manual customer success agent in some batches and none in others, to isolate what a human agent could actually catch and fix that a standard ticketing system could not.
The test surfaced concrete patterns in student pain points around product navigation and course access. Ujjawal used those findings to build the case for scaling manual support significantly, and the team grew customer success from a small unit to more than 200 agents.
30-day retention moved from 33% to 45% during this period. Ujjawal was direct that COVID-driven demand alone was pushing retention up across the board, so the test signal could not be fully isolated from that macro effect, but the jump coincided with the intervention and gave the team confidence to keep investing in the model.
He also scoped the next phase: a roadmap to automate parts of the manual support workflow using WhatsApp-based AI tools like Limechat and Yellow.ai, so the team could keep the responsiveness of manual agents without the headcount cost as volume kept growing.
Ownership Snapshot
Broad role
Headed customer experience and retention strategy reporting to the COO.
Goal
Improve student retention during a sudden COVID-driven demand surge.
Direct ownership
Designed and ran the A/B test comparing manual customer success agents to a ticketing system.
Team execution
Worked with product and customer success teams to scale the support model that resulted from the test.
