Case study · SHOPSY (FLIPKART) · 2023

Ratings & Reviews Solicitation

A contextual solicitation flow that made leaving a review effortless, lifting the share of products with ratings & reviews by 11pp.

Role
Product Designer
Team
Me · 1 PM · 2 Engineers
Platform
Shopsy App

On Shopsy, ratings & reviews are the single biggest driver of purchase confidence. Shoppers rarely buy a product with no reviews. Yet a large chunk of the catalogue, especially new and long‑tail products, sat in a cold‑start state with no reviews at all, suppressing their conversion and discoverability.

Users bought products on Shopsy but almost no one came back to review. The existing touchpoint was on the Order Details Page which was missed by users. Other problems included:

  • Bad timing. Review prompts popped-up when intent was lowest, far from the moment the product was actually in the customer's hands.
  • Too much effort. Writing a full review from a blank box was heavy; most people bounced there.
  • Cold‑start drag. Products with zero reviews converted worse and got less surface area, so they stayed reviewless.
A teardown of Flipkart's legacy review flow across four annotated screens: an Order Details page where the star rating doesn't look clickable and there's no obvious place to add photos, a Review Product screen with rectangular images, a rating that can't be changed and a small bottom-left Add Image CTA, a screen forcing users to manually type everything into an unrealistically long text box, and a Thank You screen that says 'you are awesome' even for a bad rating, with a not-thumb-friendly list of earlier purchases.
Teardown of the legacy review flow: bad timing, a heavy free-text ask, hard-to-find photo upload, and tone-deaf confirmation copy.
  • Helpful, not nagging. Ask too often and it becomes spam; ask too rarely and coverage stalls. I tuned frequency and timing rules with the team to stay on the right side of that line.
  • Quality vs. volume. Lowering effort risks shallow reviews. Structured tags let us keep submissions fast while still capturing decisions‑useful signals.
  • Fraud and trust. Easier input invites spam, so the flow had to work without adding friction for genuine buyers.

I led the end‑to‑end solicitation flow, from the trigger logic to the input model. The current prompt came at the wrong place/time and demanded too much. I explored multiple approaches to make reviews seamless. Having a single‑step star tap and progressive disclosure captured far more signal without scaring people off.

  • Right moment, right place. Ratings modal surfaced just after delivery, when the product was fresh in hand. We explored multiple touchpoints like Homepage, Notifications, Order Details Page, etc.
  • One tap to start. A star rating is a single tap; a written review and photos are optional add‑ons, never a mandate.
  • Structured tags. Quick‑pick attribute tags (fit, quality, value) let people contribute useful signals in seconds, even with no words.
The Shopsy post-delivery rating flow across four screens: a 48x48 thumbnail prompt with a one-tap star rating and Remind me later, an expanded sheet with star-with-emoji rating, aspect pills (Fabric, Style, Comfort) and a large Add Photos CTA with social proof, the same sheet with photos added and tags selected, and a confirmation screen with copy based on the rating plus prompts to rate earlier purchases.
The end‑to‑end solicitation flow: a lightweight one‑tap star prompt that progressively reveals optional tags, photos, and earlier purchases — never gating the first rating.

The contextual solicitation flow lifted the share of products with ratings & reviews by roughly 11pp and images reviews solicitation almost doubled. Newly‑reviewed products converted better, and the long tail of "no reviews yet" listings reduced.

What I'd do differently: I treated images as a secondary add‑on, but in retrospect image reviews drove the most downstream conversion of all. If I ran it again I'd invest a little more in nudging to add images in the same flow or remind users later.

Want the full story? The deck walks through the whole process including research, data analysis, design explorations and final UI.