An evidence-led evaluation of the Amazon India desktop web journey from cart through order confirmation and cancellation. The redesign concepts aim to make costs, choices, and recovery easier to understand.
Download editable PowerPoint presentation (.pptx) · Download nine editable checkout designs (.pptx)
Independent educational UX case study. Not affiliated with or endorsed by Amazon. Seven genuine desktop-web screenshots were captured on 20 October 2026. One cash-on-delivery order was placed and immediately cancelled to document the confirmation and cancellation states. Personal address and payment details are excluded from the public images. No usability outcome was measured.
Observed web path: shopping cart → optional “Need anything else?” screen → payment selection, with a link to address selection → order review with delivery options. The cash-on-delivery order was placed and immediately cancelled; both resulting states were captured. The objective is to identify friction, propose clearer patterns, then test whether those patterns help shoppers complete a simulated purchase confidently.
Captured the signed-in Amazon India desktop website in Edge on 20 October 2026, following the flow through order confirmation and cancellation. Public crops exclude personal details; five display copies also remove only the browser cursor. Review each observation against Nielsen’s heuristics and relevant interaction laws. Follow with moderated usability sessions to test the redesign.
Captured in Edge on 20 October 2026. The images are cropped to exclude personal account, address and payment details. Stage order reflects this session; Amazon may vary checkout by item, account and payment method. Cash on delivery was selected; the order was cancelled immediately, before any payment or shipment.
Five display crops have only browser cursor artifacts removed; the original captures are retained in the evidence folder. The screenshots establish visible checkout states, not a usability failure or measured outcome.
Severity uses Nielsen’s 0–4 scale. Scores are provisional triage estimates until screen evidence and user testing support them.
Heuristic: Visibility of system status. Observed evidence: the review screenshot shows a ₹7 cash on delivery fee and ₹214 total; whether shoppers notice the change needs testing. Potential impact: uncertainty or abandonment at review. Proposed redesign: persistent, itemized summary with a clearly labeled final total.
Heuristic: Aesthetic and minimalist design. Observed evidence: the cart presents coupon and promotional choices; checkout also showed a “Need anything else?” screen. Their effect on attention needs testing. Potential impact: longer decision time. Proposed redesign: group choices and reveal optional details on demand.
Heuristic: User control and freedom. Observed evidence: the review page exposes Change links for address and payment; preservation of choices after editing needs testing. Potential impact: rework or accidental orders. Proposed redesign: contextual Edit links that retain state.
Heuristic: Match between system and real world. Observed evidence: card, net banking, UPI scan and pay, and cash on delivery labels are visible; comprehension needs testing. Potential impact: hesitation or method selection errors. Proposed redesign: plain-language labels with concise explanations.
Heuristic: Help users recognize, diagnose and recover from errors. Evidence needed: a naturally encountered payment failure and recovery path; neither was captured. Potential impact: duplicate payment concern and lost orders. Proposed redesign: explicit transaction status, safe retry, and change-method actions.
Heuristic: Flexibility and efficiency of use. Evidence needed: target measurements and task observation. The cart and review captures show the controls. Potential impact: missed clicks or slower completion. Proposed redesign: prominent action labels and adequate click areas.
More and more complex choices can increase decision time. For UX-02, group delivery and payment options, then disclose secondary detail as needed. Expected benefit: faster choice. A violation has not been verified.
Small or distant targets are harder to acquire. For UX-06, give the primary action and Edit controls clear clickable areas. Expected benefit: fewer missed clicks. Actual targets still need measurement.
Shoppers bring expectations from familiar checkout patterns. For UX-03, preserve choices and offer direct Edit controls on review. Expected benefit: more predictable navigation. This is a design consideration pending evidence.
“Before” uses genuine Amazon desktop screenshots. “After” is a proposed editable design concept. The recovery comparison has no original failure evidence.
Total, fees or discounts may need extra inspection before committing.
An always-readable order summary shows subtotal, delivery, savings and final payable amount. Why: it supports cost verification at the decision point.
Changing address or payment may require navigating away from review.
Contextual Edit controls open the relevant step and preserve earlier selections. Why: it reduces rework.
A payment failure may leave transaction status unclear.
A clear failed state explains whether payment was taken and offers safe retry or change method. Why: it supports recovery and trust.
| P1 · Payment recovery | Highest potential trust and transaction risk; validate UX-05 with an observed failure flow before implementation. |
|---|---|
| P1 · Price clarity | Check whether the total is understandable at review; test an itemized summary. |
| P2 · Editing and choices | Test preserved state, grouping and terminology with first-time shoppers. |
| P3 · Control polish | Measure target size and task performance, then refine placement and labels. |
Recruit five India-based desktop web shoppers with a mix of checkout familiarity and payment preferences. Ask each to update quantity, change address, choose delivery, review cost, choose a simulated payment method, recover from a simulated failure, and identify the confirmation state. Observe hesitation, wrong turns, missed controls, cost comprehension and recovery confidence. A task succeeds when the participant reaches the intended state without moderator help.
Limitations: small qualitative sample, simulated payment, no live Amazon transaction, no measured baseline. This plan has not yet been run. The current browser walkthrough is interface evidence, not a participant usability test.
| Completion / abandonment | Share of started checkouts that complete / leave; baseline not measured. |
|---|---|
| Time and interactions | Elapsed time and clicks to completion; baseline not measured. |
| Failure recovery | Share who resolve a simulated payment failure; baseline not measured. |
| Clarity and task success | Post-task confidence rating and unaided task completion; baseline not measured. |
A useful UX audit separates observation from interpretation. Screenshot evidence should determine which issues are real; a prototype makes recommendations testable; task results should decide what to change next. The present concepts are ready for validation, not claims of improved checkout performance.