projects.app - search.revamp
A comprehensive redesign of the core search experience to improve discoverability, relevancy, and user decision-making for a major healthcare platform.
// PROJECT OVERVIEW
60%
of users landing on the Consult homepage went directly to Search — making it a primary entry point and decision-making surface, not a secondary utility.
We initiated Search 2.0 after this insight surfaced. A detailed search analysis followed to understand exactly where the experience was breaking down for users trying to find the right doctor, hospital, or speciality.
// KEY INSIGHTS FROM ANALYSIS
Low-value results ranking high
Clinics with only 0-2 doctors appeared at the top, pushing relevant options lower.
Unclear speciality differentiation
Users struggled to understand the difference between multiple speciality options.
Missing symptom mapping
Queries like "skin" or "gas" showed specialities but not related symptoms, breaking the mental model.
Insufficient doctor context
Location and experience, the critical decision-making details, weren't visible in results.
Poor location relevance
Users in Gurugram were seeing results from Hyderabad or Bangalore, reducing trust.
// OBJECTIVE
✓ Improving result relevance and ranking logic
✓ Adding clear contextual information to aid decision-making
✓ Enhancing query understanding — specialities, symptoms, common terms
✓ Ensuring location-aware suggestions
✓ Creating a more intuitive, confidence-building search experience
// FLOW CHART
// THE "BEFORE" — FAILED EXPLORATIONS
Early iterations across Symptoms, Hospitals, and Procedures revealed two major UX flaws in testing:
Failed highlights
The most relevant, top-tier suggestions used a basic grey background and generic magnifying glass icon — blending into the list instead of standing out.
Fragmented visual language
Specialities, Doctors, and Hospitals each got different visual treatment and data density — forcing users to re-learn how to read results with every scroll.
// THE "AFTER" — STANDARDIZED SOLUTION
Early iterations across Symptoms, Hospitals, and Procedures revealed two major UX flaws in testing:
A unified card architecture
One standardized card format scales across all entities — Doctor, Hospital, Speciality, or Procedure now share identical visual treatment.
Clear information hierarchy
Top results elevated into distinct standalone cards, plus top-level chips (All, Speciality, Doctor, Hospital) to instantly filter what matters.
Progressive disclosure
Secondary data like booking buttons and pricing removed from search level — focus stays on Who (name), What (speciality), Where (distance).
// THE UNIFIED CARD
The same card format now scales cleanly across Speciality, Hospital, Doctor, Symptom, and Procedure results.
// LANDING PAGE
BEFORE
Users with no past searches saw a blank screen with only the search bar at the top — zero guidance, zero starting point.
AFTER
Now surfaces relevant suggestions based on location, preferences across verticals and previous visits to Apollo Hospital — useful from the first interaction.
// DESIGNS
// METRICS & OUTCOMES
+22%
Increase in doctor card clicks - better ranking + key details
~18%
Reduction in search refinements - clearer cues, fewer attempts
+10%
Increase in consult conversions from the search funnel
+30%
Improvement in location relevance accuracy
+10-12%
Increase in broad symptom queries — "skin", "gas", etc.
1
Unified card system, replacing 3+ fragmented layouts
Search 2.0 is live on Apollo 24/7 today. Future scope includes deeper symptom-to-speciality mapping, building on the broad query increases already seen.
Other Case Studies
A high-priority feature designed to facilitate rapid, on-demand medical consultations, connecting users with healthcare professionals instantly when time is of the essence.
View
FriendZone's host ranking was invisible — hosts had no idea why some earned more than others. I designed a system to make that ranking transparent, actionable, and motivating for a non-tech-savvy audience.
View