Loyalty & Referral
CMP had strong inbound but weak retention. Users ordered once and left. I framed the growth problem, defined the system, and designed a behavioural loop that turns every completed order into the start of the next one.
Team
Uday, PD
Akash, PD
Ram, Founder
Suraj, Dev
Umag, Dev
Hardik, BA
Timeline
Dec 2025 - Jan 2026
Context
About CMP
Craft My Plate was growing with new orders, strong inbound, steady momentum. But every order cost us to acquire, and most users never came back for a second one. The product ended at "Delivered."
We were building for acquisition when we needed to be building for compounding. Loyalty, referral, and feedback existed as isolated features, none of it connected. This version turns them into a single system.
Progress needs to feel real
The first decision was structural, what kind of system actually changes behaviour. Points create cognitive load. Flat discounts don’t build habit. I chose a tier model (Bronze, Silver, Gold) as the visible progress drives return behaviour.
Previous tiers were too complex. I simplified them so value is clear in one glance with current status, what unlocks next, all visible on the main screen.
Current tier
BRONZE
Upgraded to Silver
You’ve upgraded to a higher tier
300 coins
Sign up bonus
Tier benefits
View all benefits
7%
Order rewards
40%
Coins usage
Tier
GOLD
Unlock Gold tier
Maximum rewards & usage
₹50,000
more to unlock Gold tier
5000 coins
Tier bonus
Tier benefits
View all benefits
12%
Order rewards
100%
Coins usage
Tier comparison
The loop starts at delivery
Timing was the core constraint. The old feedback loop ran 4–5 days post delivery via calls as it’s too late to act. The V1 flow was worse: vague inputs, ~10 responses in 3 months, no signal. I rebuilt it with a trigger-based system. Feedback starts immediately after delivery, structured across key dimensions, each mapped to a clear action.
Low ratings route to support instantly. Weekly reports go to ops and sales. Satisfaction improved week over week. Play Store rating moved from 3.4 → 4.2 by capturing the silent majority of good exp users.
Order feedback with incentive
Feedback covers overall experience, food, packaging, delivery, and app experience. Low ratings route to the support team in the same session, issues resolved before the user churns.
App feedback
Referral timed to emotion
Referrals weren’t just placed as a static “Refer & Earn” entry. I prioritised timing and triggering referrals after a positive rating, when users are most satisfied and likely to share. The entry point still exists in the loyalty section for access, but the primary driver is contextual (major).
Tracking was unclear earlier. I made the journey visible. I also removed the separate “Submit a Lead” flow due to low ROI and merged it into referrals, reducing complexity and eliminating a redundant path.
The tracker shows every stage - invited, signed up, first order placed. Both sides rewarded on conversion. Two taps to share, no navigation required.
Guest feedback (indirect referrals)
Splash screen
Coins that mean something at checkout
The coin redemption caps were the most debated decision internally. The question was whether to make redemption uniform or tier-based. Uniform is simpler. Tier-based is more powerful, it makes progression tangible at purchase. The tier benefit is felt at checkout, not shown on a benefits page nobody reads.
One constraint was withdrawal abuse - a non-target persona was exploiting coin withdrawals, causing losses without return. I introduced eligibility criteria to restrict misuse while keeping value for the right users. Referral ROI remained strong, while abuse dropped with just a strategic change.
Cart with loyalty redeem
Withdraw coins (Incentive)
Redemption caps scale with tier - 40%, 70%, 100%. Progression is felt at checkout, not just shown on a benefits page.
Impact
Designed a post order engagement loop, supported by a lifecycle notification system, to convert completed order into repeat behaviour and referral driven growth.
+11.18%
Repeat orders (60-day cohort)
4.2
App rating has improved from 3.4
25.19%
Reward redemption at checkout
8%
users actively referred (organic acq)
37.5%
feedback completion rate
Led the end to end design of the notification system, aligning triggers, timing, and messaging across the user journey to reinforce engagement at key moments. Shifted the experience from a one-time transaction to a repeatable growth system.
Learnings
Timings is the referral feature. The same ask at the wrong moment converts at a fraction of the rate. Coins aren’t discounts, framing it as earned rewards, not cash back, changes how users relate to them. Feedback is infrastructure system , without structured signal, the reward system has nothing trustworthy to build on. Progress visibility drives behaviour. Users chase the next tier once they can see how close they are.
My role
I led end-to-end design, from framing the growth problem to shipping the full system. Strategy, flows, copy, edge cases, handoff. Worked directly with the founder and cross-functional team to define tradeoffs and ship within operational constraints.
The right product hire is a leverage decision. Let's see if that's me.
Latest update - April 2026 | All rights unreserved