A Case Study
The ₹500 Question
Millions ask the same ₹500 question. The answer can unlock Meesho’s next wave of trust.
TL;DR
Users asking “Where’s my refund?” is a top driver of avoidable support tickets and lost trust.
A simple tracker that shows every step, gives clear time promises, and sends updates before they ask.
≥20%
reduction in ‘Where’s my refund?’ tickets.
KPI = Key Performance Indicator, our main metric for success.
This is a simple info screen using data we already have. No heavy engineering lift.
₹5–₹80 Cr/year
Estimated annual savings.
Modelled range; see ROI for assumptions.
Support is 95% automated; proactive clarity is the next efficiency win.
What Our Customers Are Saying
Snapshot of public Voice of Customer (VOC). Themes cluster around delays, opacity, and refund credibility.
📦 Pickup / Return Delay
Multiple posts · 2023–2025
“Pickup postponed/cancelled multiple times… still struggling to get my money back.”
💸 Marked Refunded, Not Received
Recent · Aug–Sep 2025
“App shows ‘refund successful’—but nothing in my bank account.”
🌀 Opaque Status / Copy
Ongoing
“Return request 11 days back… still no… staff has no clue.”
🔐 Security / Fraud Fear
Mixed · 2023–2025
“Got a ‘refund help’ call with a link—looked official. Was it a scam?”
Note: Public forums over-index on negative experiences and are not representative samples. We use them to surface failure modes and then validate internally with ticket analytics and call transcripts.
The User: Meet Priya
Priya is a 28-year-old homemaker in Lucknow. For her, a missing ₹500 refund isn’t a “minor delay” — it decides if she buys groceries today or waits. When her refund is in limbo, she doesn’t think about payment gateways. She thinks: “Did I just get scammed?”
Multiply Priya by millions, and you see why ₹500 isn’t a small number — it’s a systemic trust issue.
This isn’t a “high ticket volume” problem; it’s a human trust problem.
The Solution: A Radically Simple Tracker
We need to show users the wheels are turning. Each step will show a plain-English time estimate and what to expect next. We’ll also send proactive notifications as the refund moves from one stage to the next.
Orders
Refund tracker
Bedsheet Set
Order #M-023-88
₹699 credited • Sep 02
Visualizing the Refund Journey
1
Pickup Scheduled
2
Item Picked Up
3
Checked & Approved
4
Refund Sent to Bank
✓
Amount Credited
The Plan: Start Small
This is our Minimum Viable Product (MVP): the smallest build that delivers real user value.
✓ What We’ll Build Now
- A simple info screen (engineers call this read-only).
- Use data we already have from other systems.
- Track clicks and views to see if people are using it.
- Show “usually takes X days” estimates for each step.
🚫 What We’ll Skip For Now
- No real-time GPS tracking of the delivery courier.
- Users cannot change where the refund goes.
No live chat on this screen (they can still go to the help center).
How We’ll Know It’s Working
Who We’ll Test On
First-time returners in 4–6 cities, split 50/50 into two groups.
The Main Goal
Reduce refund-related tickets by ≥20%.
Other Things We’ll Watch
Do they order again sooner? Are they happier with the refund process?
Ensuring a Fair Test
We’ll pause other refund-related tests to make sure we’re measuring only the impact of this change.
This is called attribution—making sure credit goes to the right cause.
How This Saves Money (ROI Model)
Savings ≈ C × 365 × r × d × u
Methods note: ticket volume and ROI inputs are modeled from public statements and industry ranges.
~100k
Daily Tickets (C)
5–15%
Refund Share (r)
20–35%
Tickets Prevented (d)
₹3–10
Cost Per Ticket (u)
₹5–₹80 Cr/year
Modelled savings band.
Even the low end (₹5Cr) pays back in under three months.
Even if our assumptions are off by half, the ROI pays back in under six months.
What Could Go Wrong (Risks)
Delays from partners could make the tracker look dead, killing trust.
Show “usually” estimates and clearly flag any delays to the user.
Corporate jargon (‘Refund Initiated’) kills trust.
Use plain English and always explain what happens next.
Bank “last mile” feels like a black box and spikes anxiety.
Show specific time promises for each payment type (e.g., UPI vs. cards).
Showing a delay might actually cause more people to contact support.
Offer a one-tap “remind me” option instead of a “contact support” button.
Sources
- ¹ Meesho scale: Business of Fashion, ETtech.
- ² Support automation: Business Standard.
- ³ Public VOC: Trustpilot, Sitejabber.
- ⁴ Amazon refund help: Status, Timelines.
- ⁵ Flipkart refund policy: Policy, Guide.
- ⁶ Meesho refund window: WareIQ.
- ⁷ GenAI voice bot savings: TechCrunch.