Skip to content
Case study

Flipkart: the return that costs twice

A slow refund is annoying. A silent one is what actually breaks trust.

Reframes 'reduce returns' into 'make refund progress visible', and proposes a North Star built around trust, not speed.

Root-cause framingJourney mappingCompetitor analysis
Flipkart: the return that costs twice full deckOpen full screen

What it is

The wait is worse than the delay.

You hit return on Flipkart, and then you wait in a void: no ETA, no pickup notice, no one to call. This teardown's finding is that returns break trust through silence, not speed — and the fix that matters most is the cheapest one. A team teardown; I led problem discovery and the Amazon competitor analysis.

Returns are not an edge case in India

25%
of online apparel orders come back
16.5%
the global average, for contrast
20–65%
of item value eaten per return

Where trust breaks

The same return, two experiences

Flipkart today
  1. The return button hides — users land in Help before they've even started.

  2. Pickup happens unannounced, with no contact and no slot.

  3. The refund goes silent: no progress, nothing to watch.

  4. The user assumes the worst and contacts support.

What Amazon does
  1. Honest policy up front, no surprise exclusions.

  2. The user picks the pickup slot.

  3. Refund status goes live the moment pickup completes.

  4. Findable support, with a call-me-back option.

Tradeoffs

Where I'd spend, and where I wouldn't

Chose

Ship refund-status visibility first

over Building open-box instant returns

Visibility lives in the UI and policy layer, needs no new logistics spend, and directly attacks the trust gap. Instant returns feel impressive but cost real money at 20 to 65% of item value per return; they come later, once the cheap fix is proven.

Chose

Hold sellers accountable for 'not as described' returns

over Penalizing the buyer

These returns are caused by the supply side. Routing the cost and the fix to the seller corrects the actual source; punishing buyers corrects nothing and erodes the trust you're trying to protect.

The metric that matters: returns completed without the user ever asking where their refund is.
North Star — refund confidence rate
Next · case study

Zepto: adding to an order that's already on its way

A teardown of the moment you remember one more thing after checkout.

Continue