yoyo

yoyo is a mobile shopping app for impulse buys, and this PRD defines a mobile checkout flow that compresses the process from five screens to three. It helps frequent mobile shoppers aged 18-35 who represent 65% of our traffic, by cutting completion time by 50% to boost conversions. This matters now because mobile abandonment rates hit 72% last quarter amid rising iOS 17 adoption and competitor flows like Amazon's one-tap, risking yoyo's 15% market share if we don't ship fast.

Shoppers on yoyo abandon 72% of carts during mobile checkout due to mandatory logins, repetitive data entry, and slow payment loading that averages 45 seconds per step—worse on LTE connections. User surveys from 500 respondents show 61% cite "too many fields" as the top frustration, with session recordings revealing 40% drop-off right after address entry. This directly tanks revenue, as abandoned carts represent $2.5M in lost Q3 sales, and competitors like Shopify apps convert 28% better with saved defaults.

As a returning shopper, I want the checkout to auto-populate my saved shipping address and payment method so I can complete the purchase in under 90 seconds without re-entering details.

• System loads the most recent valid address and default payment on flow start.
• User can edit fields inline; validation happens on blur with clear error messages like "ZIP code required."
• If no saved data exists, prompt shows but doesn't block; flow proceeds to manual entry.

As a guest user, I want an option to skip login during checkout so I can buy without creating an account.

• Checkbox appears on first screen: "Continue as guest"; selecting it hides login fields.
• Guest flow collects email for receipts but skips password; order history requires later signup.
• On completion, upsell login with "Save for next time?" modal, but dismissible without penalty.

As any shopper, I want payment selection to support Apple Pay and one saved card natively so I avoid typing card details on mobile keyboards.

• Display saved cards and Apple Pay button first; tapping proceeds directly to confirmation.
• For new cards, capture via secure form with CVV mask; reject incomplete inputs before submission.
• Confirmation screen recaps total, taxes, and shipping; user must tap "Place Order" to finalize—no auto-submit.

As a shopper in a hurry, I want real-time order total updates during address selection so I know shipping costs before committing.

• Total refreshes instantly when address changes, pulling rates from backend in <500ms.
• If rates fail to load, show cached estimate with "Estimating..." spinner; proceed only if user confirms.
• Highlight changes like "$5 added for expedited" in bold to prevent surprises.

As an international shopper, I want currency and tax calculations to auto-adjust based on detected location so I see accurate pricing without manual tweaks.

• On address entry, detect country from IP or GPS (with permission); convert prices in real-time.
• Display fallback USD if detection fails, with note "Prices may vary by location—confirm address."
• Block submission if tax calculation errors exceed 5%; route to support chat instead.

• Checkout completion rate ≥ 35% (from current 18%), measured as orders placed / carts initiated in mobile sessions.
• Average time to checkout completion ≤ 90 seconds, tracked via frontend timestamps from start to order confirmation.
• Mobile cart abandonment rate ≤ 50% (down from 72%), calculated as carts viewed but not purchased per day.
• Adoption of saved payment methods ≥ 60% among returning users, via analytics on flow paths taken.
• Error rate during checkout ≤ 2%, including validation failures or network timeouts, logged per session.

• Network failure mid-flow: Cache all entered data in device storage; on reconnect, resume from last step with a 30-second timeout retry, falling back to error screen if >3 attempts. Last launch crashed 12% of flows on spotty WiFi—must test with 2G simulation.
• Permission denial for location services: Default to manual address entry; don't prompt again in session to avoid iOS blocks, but log for future onboarding nudges. Android denials spiked errors by 8% in beta.
• Low battery or app backgrounding: Pause timers and save state; on resume, warn "Session expired?" if >10 minutes idle, requiring restart to prevent stale data leaks. Previous version lost 5% of carts here.
• Invalid promo code entry: Apply only if valid; on failure, shake field with "Code expired" message, but allow proceed without it—no hard block to avoid frustration. Codes failed silently before, inflating support tickets.
• High-value orders (> $500): Require explicit CVV re-entry even for saved cards; add fraud check delay <1s, or block with "Security review needed" and manual support escalation. Fraud losses hit $50K last quarter on unchecked flows.
• Device rotation or split-screen: Maintain state across orientations; test on foldables where keyboard overlap hid submit button in 20% of cases during alpha.

• Which payment processors to prioritize beyond Apple/Google Pay—Stripe for cards or Adyen for global? Need decision before API integration. ⚠ Critical, blocks security audit.
• How to handle multi-currency display for border shoppers—auto-convert or show dual? Affects tax calc accuracy; resolve in design review next week.
• Opt-in for biometric auth (Face ID) on payment: Mandatory or toggle? Impacts conversion; user testing pending, low urgency.
• Integration with loyalty points redemption in flow—add as fourth step or inline? Depends on backend readiness; flag for PM sync. ⚠ Delays shipping if not scoped.
• iOS 18 beta compatibility: Test for widget shortcuts to checkout? Exploratory, post-launch.

• Backend team: User profile API (v2.3) for saved data fetch/update; must support 500ms p95 latency or flow degrades. Already delayed twice—escalate.
• Payments team: Stripe API keys and webhooks for tokenization; requires sandbox testing environment live by sprint start.
• Design team: Figma prototypes for three-screen flow, including dark mode variants; assets due EOW.
• Infrastructure: Feature flag via LaunchDarkly to toggle new flow for 10% users initially; iOS/Android builds need crashlytics integration for error tracking.
• Third-party: Apple Pay certification (entitlement approved) and Google Pay API (sandbox endpoints); no fallback if revoked mid-launch.
• Analytics: Amplitude events for checkout paths; depends on schema update from data team to avoid metric gaps like last failed rollout.