Emami Frank Ross Pharmacy App - Redesigned
Diagnosing a 75% Day 1 uninstall rate and redesigning the pharmacy experience end to end.
Case Study - Health and Pharma
Frank Ross had a pharmacy app that 75% of users uninstalled within Day 1. I diagnosed the behavioural and structural causes through real analytics, Play Store review mining, heuristic evaluation, and competitive research - then redesigned the end-to-end experience from home screen to post-purchase. This case study documents the discovery, decisions, and design process behind that work.
Context
The Business Problem
Emami Frank Ross Ltd is one of India's oldest pharmacy chains - over 200 physical stores across Kolkata and nearby cities, a strong brand legacy, and a loyal offline customer base. The company had built a mobile app to capture online sales and expand beyond West Bengal into cities like Bengaluru and Mysore.
But the app was not working. The data told a damning story before I wrote a single wireframe.
of users uninstalled within Day 1
customer retention rate
cart abandonment rate
search-to-conversion rate
These were not isolated design problems. They were symptoms of a product that failed to earn trust, reduce friction, or give users a reason to return. My job was to find out why - and fix it from the ground up.
Role
My Role on This Project
I was the solo Product Designer, working alongside a Project Manager and Product Manager at Tenovia. I owned the complete design process - discovery, research synthesis, information architecture, wireframes, high-fidelity UI, prototype, and developer handoff.
Before sketching anything, I ran a working alignment session with the PM to agree on scope, constraints, timeline, and what "success" meant in measurable terms. That conversation defined the KPIs that every design decision was later judged against.
Research
Discovery and Research
I approached discovery in three layers - the analytics, the product itself, and the market. Each layer either confirmed or challenged what the previous one suggested.
Layer 1 - Reading the data first
I started with the app's own analytics before forming any hypotheses. The uninstall data showed a cliff at Day 1. Users were not churning after a bad experience over time - they were leaving immediately after installation. That told me the problem was first-session clarity and trust, not feature depth or long-term engagement.
The new vs returning user trend revealed a more complex picture. New user acquisition was declining month-over-month while the returning user ratio temporarily elevated - a false positive. The product was not retaining well. It was simply failing to attract new users consistently enough to grow the denominator.
Layer 2 - Heuristic evaluation and Play Store review mining
I ran a structured heuristic evaluation of the existing app against Nielsen's 10 usability principles. What I found was not subtle. The navigation used pharmaceutical jargon - OTC, Rx - that everyday consumers could not parse. Search returned irrelevant results. Product pages were missing basic purchase-decision information: expiry dates, return policies, delivery timelines. The checkout required 5+ clicks for actions that should have taken one.
I cross-referenced all of this with Play Store and App Store reviews to validate which friction points users were actually expressing in their own words.
"I was really let down by the app - it did not meet my expectations in terms of design and user-friendliness."
Layer 3 - Competitive and secondary research
I analysed how 1mg and PharmEasy - the category leaders - handled the same user flows. Both had invested heavily in trust signals, search quality, and simplified prescription workflows. Frank Ross was competing in the same category without the same foundational UX in place.
Secondary research on Indian pharmacy app consumers consistently surfaced the same priorities: convenience, product authenticity, fast delivery, and an interface that does not require medical literacy to navigate.
Google Trends confirmed the gap - Frank Ross had strong offline brand recognition in West Bengal, but negligible digital search presence compared to Apollo Pharmacy and 1mg nationally.
Research Synthesis
User Persona and Journey Map
Based on the analytics, review mining, and competitive research, I defined the primary user - not a healthcare professional, but an everyday urban consumer who wants to order medicines from home without confusion or anxiety. They are not looking for clinical depth. They want simplicity, reassurance, and speed.
Strategy
Defining Success Before Designing
One of the first things I did - before any wireframe - was work with the PM to lock in specific, measurable KPIs. Any design decision that could not be traced back to one of these metrics needed a stronger justification. This kept design debates grounded in outcomes, not preferences.
| Metric | Baseline | Target |
|---|---|---|
| Conversion rate from search | 9% | +15% |
| Products added to cart from search | 17% | +30% |
| Items added to cart from category page | 15% | +33% |
| Average order value | INR 685 | INR 705 |
| Cart abandonment rate | 30% | -25% |
| Coupon apply rate | 40% | +75% |
| Customer retention rate | 4% | +100% |
| App uninstall rate | 75% | -10% |
Structure
Sitemap
Before touching any UI, I restructured the IA around how users think about their health needs - not how a pharmacy categorises its inventory. Categories sorted by pharmaceutical type made internal sense but failed users who think in terms of symptoms and conditions.
I relabelled all Level 1 categories in plain consumer language, sorted them by purchase frequency, and restructured the bottom navigation. The Articles link - receiving 0.5% of all clicks - was replaced with My Orders, which is critical for less tech-savvy users navigating the post-purchase experience.
Design
Design Decisions - Screen by Screen
Every screen change was tied to a specific behavioural or business problem identified in research. Here is the reasoning behind each major decision.
Home Screen
- Oversized header consumed all first-fold real estate
- Minimal content - nothing to explore or engage with after landing
- Poor category labelling - confusing for new users
- Articles link in bottom nav at 0.5% click rate - wasted prime navigation space
- Optimised header to surface offers and frequent reorders above the fold
- Added navigational sections: top deals, popular categories, shop by brand, shop by concern
- Replaced Articles with My Orders - critical for post-purchase trust and return visits
- All important links promoted to top with scannable small cards
Category List
- Jargon labels (OTC, Rx) unfamiliar to general consumers
- Categories ordered randomly, not by purchase popularity
- No thumbnails - low visual scanability and discoverability
- Relabelled all Level 1 categories in plain consumer language
- Sorted by purchase frequency - bestselling categories surface first
- Top-selling product images used as thumbnails for instant recognition
- Visual hierarchy added to show users their current position
Product List Page
- Incorrect product pricing displayed
- Out-of-stock only revealed after user taps Add to Cart - multiple times in a row
- No product ratings - a critical trust and purchase-speed signal
- No way to get notified when out-of-stock items return
- Accurate pricing with Best Deal and Best Seller tags for faster scanning
- Out-of-stock surfaced upfront with Notify Me - eliminates the dead-end loop
- Ratings added - accelerates purchase confidence without adding cognitive load
- Add to Cart simplified from 3 interactions to 1
Product Detail Page
- No expiry date, delivery timeline, return policy, or available offers
- No upselling or cross-selling - missed revenue at highest-intent moment
- No trust signals - users could not confidently commit to buying
- Product variants missing entirely
- Full product information in a consumable format - expiry, delivery, returns, offers
- Frequently Bought Together and Similar Items added to increase average order value
- Trust badges, ratings, and reviews section added to reduce purchase anxiety
- Product trend indicator added to accelerate decision-making
Cart and Checkout
- Cluttered, outdated UI with no visual hierarchy
- Coupon apply required 5 clicks - an unnecessary friction tax
- No removal confirmation - accidental deletions caused frustration
- Prescription and non-prescription items mixed without separation
- Prescribed and non-prescribed items clearly separated for clarity and compliance
- Coupon apply reduced to a single inline action - removing friction at the highest abandonment point
- Removal confirmation step prevents accidental cart emptying
- Cross-sell added at cart stage for contextual AOV uplift
Prescription Upload
- 4-5 screens for a task that should take one
- Active secondary button styled identically to a disabled state
- No upfront information about which items required prescriptions
- Entire flow collapsed into one screen - progressive disclosure without overwhelming
- Items requiring prescription surfaced at top of the list (recognition over recall)
- Upload or skip options presented clearly without creating unnecessary decision paralysis
Order History
- Poor UI with insufficient order status information
- No feedback collection mechanism post-delivery
- Cancel button too prominently placed - driving unnecessary cancellations
- No search or filter for returning users with long order histories
- One-click rating system for post-purchase feedback - lowest possible effort
- Order status highlighted with colour-coded states for instant comprehension
- Cancel button moved to detail page, placed beside Customer Support to guide better decisions
- Search and filter added to support returning users
Outcome
Final Designs and Prototype
An interactive prototype was built in Figma to simulate the real experience - micro-interactions and screen transitions included - giving the client and development team a precise picture of intended behaviour at every touchpoint.
Validation Plan
What I'd Validate Next
The design was fully completed and handed off to the client team. Post-implementation measurement was outside my project scope. If I were to validate this work, here is exactly how I'd approach measurement - in three phases, each building on the last.
Did we fix the first-session failure?
The 75% Day 1 uninstall was the most urgent problem. Everything in month one is about whether the new experience earns enough trust to keep users past that critical first session.
- Day 1 and Day 7 retention rate - primary signal that the home screen, navigation, and onboarding are no longer driving immediate abandonment
- Search-to-product click rate - is improved search returning results users actually want?
- Prescription upload completion rate - does the collapsed single-screen flow reduce drop-off at this critical friction point?
- App Store and Play Store review sentiment - qualitative early signal before quantitative data matures enough to read confidently
Are users converting and transacting?
Once retention stabilises, the focus shifts to the purchase funnel. Are users who stay actually buying? Are the friction reductions translating into revenue?
- Search-to-cart conversion rate - target was +15% from the 9% baseline; direct test of search quality and PLP improvements
- Category-to-cart conversion rate - target was +33% from 15%; validates the IA restructure and plain-language labels
- Cart abandonment rate - target was -25% from 30%; the coupon simplification and removal confirmation are the primary levers here
- Coupon apply rate - target was +75% from 40%; a direct test of whether reducing friction from 5 clicks to 1 changes behaviour at scale
- Average order value - watching for AOV movement driven by Frequently Bought Together and Similar Items on the PDP
Are we building a sustainable product?
Month three is about durability - whether the improvements hold over time and whether users are forming habits around the product.
- Customer retention rate - the +100% target (4% -> 8%+) becomes statistically readable here; this is the headline metric for the entire engagement
- Post-purchase rating submission rate - baseline for the feedback loop; if users are rating, they are engaged enough to have formed an opinion
- Uninstall rate trend - are we holding users we previously would have lost after Day 1?
- Repeat order rate - the ultimate proof that the experience is worth returning to; this is what turns a pharmacy app into a habit
The KPIs were defined upfront precisely so that any team implementing this work could measure against a clear benchmark from day one - not retrofit success criteria after the fact.
Reflection
What This Project Taught Me
Analytics before assumptions. The Day 1 uninstall cliff told me exactly where to focus before I'd spoken to a single user or drawn a single wireframe. In healthcare apps, the stakes of getting the first session wrong are higher than in most categories - trust is harder to rebuild once broken. Data is not a substitute for empathy, but it is an indispensable guide to where empathy needs to go first.
Language is a design decision, not a copy edit. Relabelling "OTC" and "Rx" to plain-language consumer categories was not wordsmithing - it was a navigation fix that changed how users could find what they needed. Every label that requires domain knowledge is a friction point for the majority of your users. In consumer health especially, plain language is a trust signal.
Defining KPIs before designing makes every decision defensible. Having specific metrics agreed upfront - with the PM and stakeholders - meant design debates were grounded in business outcomes, not personal preference. It also gave me a clear prioritisation framework: fix the things with the highest uninstall-driving impact first, optimise for conversion second, build for retention third.