Product Strategy · Blinkit
Frequent shoppers keep buying the same things and rarely try a new category
How one well-timed, relevant suggestion at checkout could widen what people buy, without slowing them down.
3,098+
customer conversations analysed
1
suggestion per checkout, accepted or ignored
An independent case study. It is not a feature shipped at Blinkit, and no experiment results are claimed.
The Problem
Habit beats discovery
What I looked at
The existing discovery surfaces, and the friction behind repeat-purchase behaviour: why people who shop often still stay inside the same familiar items.
How the thinking changed
Across 3,098+ customer conversations, my first hypothesis didn't hold up as stated. It moved toward contextual discovery: suggesting something that fits what is already in the basket.
Don't ask people to browse. Meet them in the middle of the mission they are already on.
The Idea
A contextual add-on at checkout
An AI-powered add-on that reads the basket and keeps the interaction light.
- 1
Infer the mission
From the basket, work out what the shopper is likely trying to do.
- 2
Suggest one item, with a reason
Recommend a single relevant item and say plainly why.
- 3
Accept or ignore
Either choice is easy, and checkout isn't slowed.
Proposed
Success Metrics
Growth that doesn't cost the checkout
North-star metric
The share of eligible checkouts where a new-category item is bought.
Alongside it: trial, repeat purchase and GMV, to show whether the new habit sticks and whether it adds value.
Guardrails
- Checkout time shouldn't rise.
- Cart abandonment shouldn't rise.
- Suggestions have to stay relevant.
The feature earns a wider rollout only if the north star moves and the guardrails hold.
Evidence Status
What is known, proposed and measured
- 3,098+ customer conversations analysed.
- The first hypothesis was revised after reading them.
- Contextual discovery at checkout.
- One suggestion, one reason, easy to ignore.
- No experiment has been run.
- No lift in new-category purchases is claimed.