BRICLAY
Gift your Friend
People don't invite friends, they send things. I designed a loop where you could send fresh produce to a friend, and it worked on screen. Then it broke on the rider's route.
Role
Sole product designer, and part-time picker, packer and rider
The problem
Briclay sold fresh fruit and vegetables, and you can't advertise freshness. People have to taste it. But asking a stranger to order before they've tried anything is a big ask.
People don't invite friends, they send things. Didn't actually wanted the referral flow to be transactional. So the idea was to let a customer send a small gift of produce to someone who had never ordered from us.
Sending a gift
Gifts are capped at two items, so they stay small and easy to say yes to. You can send to up to five friends, so one send can reach a few people. And it costs the sender nothing, so there's no reason to hesitate.
Receiving the gift
It starts with a link in SMS or WhatsApp that takes the friend to the app to claim the gift. Signing up is lighter than usual: the sender has already given us their name, so they only add a phone number and an address for the rider. Unclaimed gifts got a couple of reminders and expired after a month, so a gift never became an open-ended cost.
Trade-offs made
Every choice here bought something and cost something. These are the ones that mattered.
Decision
Why
What it cost
Two items at most
A call by the founders, to keep a gift small and cheap.
A modest gift. Nobody can send a basket and making it easier for the rider.
Up to five friends
After sending, the last screen asks you to tell them, mostly over WhatsApp, which had a limit of five people.
Someone with a bigger circle had to stop at five.
Claim in the app
The main goal was to get new customers to download Briclay. I looked at a lighter claim and dropped it.
Someone with a bigger circle had to stop at five.
Edge cases
A feature like this breaks in the corners, so designed those too.




The card that came with it
The card sat on top of the gifts inside the box, so it was the first thing a new customer saw. A box from someone they've never ordered from can feel cold, so I made the card say hello from our side.
Where it broke
The gift itself was cheap, about ₹50 of produce. The cost was the delivery. Picture a rider already on the way to a customer with their weekly order. That customer sends a gift to a friend in another part of the city, so the rider has to go there too, just to drop one or two free items. Riders were paid per delivery, so that extra trip made each new customer too expensive to win.
We saw the extra trip as a risk before launch and planned to watch it. It was removed after I left, and it's not in the app today. That's my understanding from working in ops, not an official company line.
What I'd try now
These are ideas, not shipped work. The cleanest fix is to let the gift ride along with the friend's first order, as a voucher instead of a separate drop, so no rider makes an extra trip. It would feel more restrictive as a gift, because the friend has to place an order to get it. But with the operational limits, I think that trade-off would have made sense. If it had to stay a free drop, I'd only send it on route slots that already pass near the friend, and cap gifts per area per slot.
I'd judge any of them by the claim rate, how many friends place a first paid order, and what each new customer costs, including rider pay.
What I take from it
A growth feature has two sides: the screens a customer sees, and the work someone has to do to make them true. Gift a Friend was good on the first and expensive on the second, because every gift was a rider trip. We saw the risk and planned to watch it. Next time I'd start from the cost of one delivered gift and design backwards, so the cheapest way to reach the friend ships first.
BRICLAY
Gift your Friend
People don't invite friends, they send things. I designed a loop where you could send fresh produce to a friend, and it worked on screen. Then it broke on the rider's route.
Role
Sole product designer, and part-time picker, packer and rider
The problem
Briclay sold fresh fruit and vegetables, and you can't advertise freshness. People have to taste it. But asking a stranger to order before they've tried anything is a big ask.
People don't invite friends, they send things. Didn't actually wanted the referral flow to be transactional. So the idea was to let a customer send a small gift of produce to someone who had never ordered from us.
Sending a gift
Gifts are capped at two items, so they stay small and easy to say yes to. You can send to up to five friends, so one send can reach a few people. And it costs the sender nothing, so there's no reason to hesitate.
Receiving the gift
It starts with a link in SMS or WhatsApp that takes the friend to the app to claim the gift. Signing up is lighter than usual: the sender has already given us their name, so they only add a phone number and an address for the rider. Unclaimed gifts got a couple of reminders and expired after a month, so a gift never became an open-ended cost.
Trade-offs made
Every choice here bought something and cost something. These are the ones that mattered.

Edge cases
A feature like this breaks in the corners, so designed those too.




The card that came with it
The card sat on top of the gifts inside the box, so it was the first thing a new customer saw. A box from someone they've never ordered from can feel cold, so I made the card say hello from our side.
Where it broke
The gift itself was cheap, about ₹50 of produce. The cost was the delivery. Picture a rider already on the way to a customer with their weekly order. That customer sends a gift to a friend in another part of the city, so the rider has to go there too, just to drop one or two free items. Riders were paid per delivery, so that extra trip made each new customer too expensive to win.
We saw the extra trip as a risk before launch and planned to watch it. It was removed after I left, and it's not in the app today. That's my understanding from working in ops, not an official company line.
What I'd try now
These are ideas, not shipped work. The cleanest fix is to let the gift ride along with the friend's first order, as a voucher instead of a separate drop, so no rider makes an extra trip. It would feel more restrictive as a gift, because the friend has to place an order to get it. But with the operational limits, I think that trade-off would have made sense. If it had to stay a free drop, I'd only send it on route slots that already pass near the friend, and cap gifts per area per slot.
I'd judge any of them by the claim rate, how many friends place a first paid order, and what each new customer costs, including rider pay.
What I take from it
A growth feature has two sides: the screens a customer sees, and the work someone has to do to make them true. Gift a Friend was good on the first and expensive on the second, because every gift was a rider trip. We saw the risk and planned to watch it. Next time I'd start from the cost of one delivered gift and design backwards, so the cheapest way to reach the friend ships first.