// 02 · E-commerce · Conversion
Rebuilding checkout to unlock bottom-funnel growth
Nykaa · 2022 — 2026
// Overview
At Nykaa, 70% of repeat customers don't change their delivery address and 40% don't change their payment method — yet every order forced them through the same long checkout flow.
We bet that a faster, pre-filled path for returning users would lift conversion without breaking trust. After three iterations, Express Checkout delivered ₹44 Cr topline impact, with a clear runway to ₹60 Cr+.
// Problem & context
Repeat buyers are the backbone of Nykaa's volume. But our checkout treated them like first-time users — same long page, same decisions, same friction. The data was telling us something we weren't acting on: ~70% stick to the same delivery address, and ~40% use the same payment method.
The insight wasn't that people repeat — it was that we were charging them decision tax for behaviour they'd already locked in.
The question I took to the team:
If most returning users have already made these decisions, why are we asking them to make them again?
// Data
A bottom sheet experience for repeat users that pre-fills last-used address and payment method, surfaces savings, and lets them confirm in one motion — while keeping change options one tap away.
Not a redesign. A shortcut. A fast lane that respected returning users' time without taking control away from them.
// How it evolved: three releases, one hard lesson
Each phase taught us something the data alone couldn't. And because checkout isn't a space you can observe users in real time, every iteration was a calculated bet — built on internal user testing, secondary research, and product sense.
// Phase 1: We shipped speed. Users rejected it.
Our first version — Quick Pay — was built on one hypothesis: returning users want to pay faster. It showed delivery estimate and let them pay in one tap.
It backfired. Drop-offs went up. We were puzzled.
Our hunch: users hadn't registered what the screen was asking of them. As creatures of habit, they'd tapped the CTA like any familiar button — and got surprised when the payment gateway loaded. We'd skipped the moment of recognition they needed.
Pushing harder on speed wasn't an option. The takeaway: speed without context isn't speed, it's anxiety.
// Phase 2: Context before speed
A series of educated guesses, tested internally, refined each cycle:
Replaced “Quick Pay” CTA with the final payable amount and total savings — users needed to register what they were paying before they paid.
Swapped tap for slide-to-pay (inspired by Swiggy) — a small, deliberate friction that forces the user to notice.
Questioned every word on the sheet — cut everything that wasn't the decision.
Impact:
- +240 bps uplift in Checkout → Pay Now
- +80 bps overall Checkout → Order
- −150 bps softness in Pay Now → Order (mostly OTP/PIN drop-offs)
- payment success rate 68.8% vs 72.2% in the regular flow
// Phase 3 (live): Closing the trust gap
The numbers were positive, but qualitative feedback collected from a user research flagged a softer issue: “Which payment method is this using?” “How do I change it?” “That felt too quick — I wasn't sure what just happened.”
Two changes: added a short, deliberate delay that read as progress, and visually anchored the payment method on the sheet — so users could see exactly what was about to happen.
Impact: incremental +25 bps lift, smoother progression, fewer reversals.
// What's next
Checkout isn't a space you can observe users in real time — there's no usability lab moment mid-payment. Every next move is designed on secondary research, internal testing and product judgment, which means the calls on what ships next sit with the design team.
The roadmap: motion and micro-delight in the confirmation moment, smarter pre-fill logic using behavioural signals rather than last-used, an OTP/PIN redesign to recover the −150 bps drop-off, and deeper personalisation for high-intent users.
// The bigger questions
The work also opened up larger questions I'm pushing with stakeholders. What's actually stopping us from a one-page checkout? How do we surface the right payment offer on top — automatically? Why have a payment page at all? Can users pay directly from cart?
// Outcomes
₹44 Cr in topline impact, with every 1% improvement worth roughly ₹60 Cr more.
Taken together, these numbers validated the bet. After the unexpected failure of Phase 1, Phases 2 and 3 confirmed that Express Checkout wasn't just a successful experiment — it was a permanent shift in how returning users would experience checkout at Nykaa.
// Next up