Why Zomato's checkout matters
Zomato processes millions of orders daily across India. Their checkout is the final — and most financially critical — step in every transaction. A 5% improvement in checkout conversion at Zomato's scale would translate to tens of thousands of additional orders per day.
This audit covers the mobile checkout flow from cart review to order confirmation — approximately 4 screens that determine whether a user completes their purchase or abandons. I audited the Android app (v17.x) across 3 devices in November 2024, using Nielsen's 10 usability heuristics as the primary framework.
Disclaimer: This is an independent audit. I have no affiliation with Zomato. All observations are based on publicly available app behaviour. The goal is constructive critique, not brand criticism.
10 heuristics. One honest score.
Each heuristic scored 1–10. Scores below 5 indicate significant usability issues that likely impact conversion. Scores of 4 or below are critical.
Order status is clear post-checkout but payment state feedback during processing is poor.
Language is mostly familiar. 'Pro' tier naming is clear but pricing logic uses internal jargon.
Editing order after checkout is buried. Cancellation window is unclear until it's too late.
Button styles are inconsistent between checkout and post-order screens.
No address confirmation before payment. No warning when ordering from a restaurant near closing time.
Saved addresses and past orders are surfaced well. Payment methods less so.
Power users have no fast-reorder shortcut from the home screen. Every repeat order requires full checkout.
Checkout screen has 11 distinct UI zones competing for attention. Classic case of feature accumulation without hierarchy.
Payment failure messages are generic. No specific guidance on why a card failed or what to try next.
Help is accessible but not contextual — same generic FAQ regardless of where you are in the flow.
Seven points where users drop
These are the specific moments in the checkout flow where friction is highest — ranked by estimated conversion impact.
F1 — UPI buried below fold on payment screen CRITICAL
UPI is India's dominant payment method — used by 70%+ of digital transactions. Yet Zomato's payment screen shows credit/debit cards first, with UPI requiring a scroll. This contradicts user mental models and adds unnecessary cognitive load at the highest-anxiety moment. Estimated impact: 8–12% of payment abandonment attributable to this ordering.
F2 — No address confirmation before payment CRITICAL
The delivery address is shown at the top of checkout but cannot be edited from the payment screen — users must go back, losing their coupon application in the process. This creates a fear of commitment: users scroll back to verify address, disrupting the forward flow. A persistent address chip with inline edit would eliminate this entirely.
F3 — Coupon field resets on back navigation CRITICAL
If a user applies a coupon, then navigates back to change a cart item, the coupon is cleared on return. This is a direct loss — users who applied a discount and had it silently removed are unlikely to reapply it and more likely to abandon. State should persist across the checkout session unconditionally.
F4 — Payment failure message: "Something went wrong" CRITICAL
When a payment fails, users receive a generic error with no information about why it failed or what to try next. This is particularly damaging on UPI — where failures can happen for 6 different reasons, each requiring a different recovery action. A specific error ("UPI PIN incorrect — try again" vs "Daily UPI limit reached — use card instead") would recover a significant proportion of failed transactions.
F5 — Delivery time estimate not shown in cart MODERATE
The delivery time estimate disappears from the restaurant listing when users enter checkout. Users have to remember it — or go back to check. Delivery time is a key variable in the purchase decision (ordering for lunch vs dinner has different tolerances). Surfacing it persistently in the checkout header would reduce back-navigation.
F6 — Pro upsell interrupts the checkout flow MODERATE
A Zomato Pro upsell modal appears between cart and payment — stopping a user mid-transaction to pitch a subscription. This is a textbook case of friction inserted at the wrong moment. Users in checkout mode have high purchase intent that should be completed, not interrupted. The upsell converts better post-order, when users have just experienced the service they're being asked to subscribe to.
F7 — No fast-reorder from home screen LOW
Repeat orders are a significant portion of Zomato's transaction volume — same restaurant, same items, same address. There is no shortcut to reorder without going through the full flow. A "Reorder" shortcut on the home screen (showing last 3 orders with one-tap checkout) would materially improve repeat purchase frequency.
The cognitive principles being violated
Friction points don't exist in a vacuum — each one triggers a specific cognitive response that makes abandonment more likely. Understanding the mechanism points directly to the fix.
Peak-End Rule — the payment failure ruins the memory
Kahneman's Peak-End Rule: people remember an experience by its most intense moment and its ending. A payment failure — confusing, unresolved, without clear recovery — becomes the peak-negative moment. Even if the order eventually succeeds, the memory of struggling with the payment persists. Better error recovery changes what users remember about the experience.
Reactance — the Pro modal creates resistance
Psychological reactance: when people feel their freedom is threatened, they resist. A modal interrupting checkout feels coercive — the user is trying to complete a task and the product is blocking them to pitch something else. This creates negative affect toward the Pro product specifically. Users who dismiss the modal are less likely to subscribe later than those who never saw it mid-task.
Status quo bias — UPI ordering violation
Users have established mental models for checkout — the most common payment method should be first. When the interface violates this expectation (cards before UPI), users must consciously override their learned behaviour. This increases cognitive load and error probability at the worst possible moment. Matching the interface to the majority mental model is the cheapest possible conversion improvement.
Five fixes. Prioritised by impact.
These are not design concepts — they are specific, implementable changes ranked by estimated conversion impact vs engineering effort.
Promote UPI to the top of the payment list
Impact: High. Effort: Low. Reorder the payment method list to show UPI first, followed by saved UPI IDs, then cards. This is a configuration change, not a redesign. Based on industry benchmarks, payment method ordering alone accounts for 6–10% of checkout conversion variance. Estimated implementation: 1 sprint.
Persist coupon state across session
Impact: High. Effort: Medium. Store the applied coupon in session state — it should survive back-navigation, address changes, and item additions. Show a persistent "Coupon applied: SAVE100" chip throughout checkout. This is both a trust fix and a conversion fix — users who see their discount confirmed throughout are less likely to second-guess the purchase.
Specific payment failure messages
Impact: High. Effort: Medium. Map each payment failure code to a specific, actionable message. "UPI PIN was incorrect — try again" / "Your daily UPI limit has been reached — pay with card" / "Bank server timeout — retry in 30 seconds." Recovery rate on specific error messages vs generic ones averages 40% higher in payment UX research. This requires working with the payment provider to expose error codes to the front end.
Move Pro upsell to post-order confirmation
Impact: Medium. Effort: Low. Remove the Pro modal from the checkout flow entirely. Replace it with a contextual banner on the order confirmation screen: "You just saved ₹0 on delivery. Pro members save ₹X every month." The user has just completed a positive action — they're in the best possible emotional state to hear about a subscription. Conversion on post-order upsells is 2–3× higher than mid-checkout interruptions.
Persistent delivery time in checkout header
Impact: Medium. Effort: Low. Add a single line to the checkout header: "Arrives in ~35 min". This is data Zomato already has — it just needs to be carried through to the checkout view. Eliminates the back-navigation loop caused by users trying to verify delivery time before committing to payment.
What fixing this is worth
At Zomato's transaction volume, even small conversion improvements compound into significant revenue. These are directional estimates based on industry benchmarks — not Zomato's internal data, which I don't have access to.
The most important thing about this audit is not any single fix — it's the prioritisation. Companies with Zomato's scale have significant engineering costs. The 5 changes above are ordered by impact-to-effort ratio: the UPI ordering fix and the Pro upsell relocation are configuration changes that could ship in days, not months. Starting there, then validating with A/B tests before investing in the higher-effort payment error work, is the right sequencing.
This is an independent audit conducted without access to Zomato's analytics, user research, or internal data. All impact estimates are based on published industry benchmarks and analogous product research. The intent is constructive — these are the same observations a senior UX leader would make in an internal review.
