CASE_01 / GROWTH Recovering app-store trust after payment bugs hurt ratings
Turning retained, satisfied customers into visible public proof.
Context
Payment issues had pulled app ratings down, which weakened trust for new users discovering Pickup Coffee through the Play Store and App Store. Retention and in-app order feedback told a different story: regular customers were coming back, and 90% of order feedback was positive, but that satisfaction was not reaching public ratings.
The hard decision
Created a post-order rating nudge at the moment of highest confidence: after a completed purchase. The loop moved happy customers from in-app feedback to the store rating flow without interrupting checkout or asking cold users to review the app.
Constraints
- — No reliable way to predict who would leave a negative public review
- — High-retention users liked the product but were not rating it
- — The prompt had to avoid adding friction to the order journey
- — Payment bugs had made user trust fragile
Measured against
App rating improved from 4.59 to 4.74 · 5-star review mix recovered · 1-star review share came down · ~3,000 monthly store-rating intent taps
CASE_02 / EXPANSION Leading the Mexico MVP launch for Pickup Coffee's consumer app
Country expansion through a focused MVP, localization, and launch operations.
Context
Pickup Coffee already had stores and booths in Mexico, but the consumer app only worked in the Philippines. Mexico wanted the app to unlock user growth, voucher distribution, and customer behaviour visibility without inheriting every Philippines feature on day one.
The hard decision
Scoped a lean Mexico MVP on top of the existing Philippines app structure. We removed non-essential flows like delivery, localized the experience, added multilingual support, and restricted access by market through phone-number and location rules.
Constraints
- — Mexico needed speed without a separate app rebuild
- — Philippines and Mexico users had to be routed to the right market experience
- — Localization, legal, marketing, and operations requirements kept changing
- — Baristas and local teams needed training on back office, vouchers, and analytics tools
Measured against
Mexico app launch completed · Initial user traction · Store and barista readiness
CASE_03 / COST SYSTEMS Reducing SMS costs by moving order updates to push notifications
A cost spike investigation turned into a recurring savings system.
Context
SMS costs started rising without a clear owner or root cause. Marketing campaigns sent through CleverTap were part of the spike, but a deeper analysis showed that recurring order-status messages from the backend were the larger structural cost driver.
The hard decision
Split the problem into campaign SMS and transactional SMS. Marketing campaigns were capped after low conversion was identified, while order updates moved from repeated SMS messages to push notifications, with one SMS reserved to tell users they could track the order in-app.
Constraints
- — Order communication still had to feel reliable to customers
- — Marketing needed limits without losing campaign reach
- — Backend-triggered status updates touched multiple order states
- — Push notifications depended on clean user and order data from CleverTap integration
Measured against
~$3K/month SMS cost reduction · ~$36K/year savings · Marketing SMS caps introduced · Order updates shifted to push notifications
CASE_04 / ANALYTICS Building product analytics instrumentation for funnel & behaviour visibility
Measurement maturity, not vanity dashboards.
Context
Tracking was inconsistent. Naming was duplicated. Funnels could not be trusted. PM, marketing, and engineering each held a different version of the truth.
The hard decision
Designed a single event taxonomy across homepage, location, onboarding, reorder, cart, rating, and loyalty surfaces. Defined parameter conventions and a single ownership matrix so analytics stopped drifting between PRDs.
Constraints
- — Existing CleverTap events could not be retroactively renamed
- — Marketing needed campaign-level attribution preserved
- — Mobile release cadence limited backfill speed
Measured against
Event coverage % · Funnel reliability · Decision turnaround time · Duplicate event reduction
