Index / Selected Work / Case 03 · Room Deduplication
Product Design UI Design Information Hierarchy Filtering

Simplifying hotel room selection.

Redesigned a room-picking experience overwhelmed by vendor duplication, using deduplication, hierarchy, and smart filters to bring it closer to how people actually shop for hotels.

✦ View detailed UI ↗
Plate 01 · Hotel rooms, after deduplication ITILITE · 2024
My Role
Lead Product
DesignerEnd-to-end ownership · Research → UI
Timeline
10 weeksWith PM, frontend & backend engineering
Users
Business
travelersBooking hotels on ITILITE, high-volume, repeat
Impact
~30 → ~6
rooms / hotelFaster discovery · Lower drop-off · Higher confidence
01 · Problem context

A room page that felt more like a warehouse than a hotel.

ITILITE aggregates hotel inventory from multiple vendors. As supply grew, the room selection experience scaled in the worst way possible, duplicated, inconsistent, and impossible to compare at a glance.

25 – 30 rooms
Ungrouped cards per hotel page
Vendor duplicates
Similar rooms repeated under different vendors
Inconsistent names
Room names & amenities varied across sources
Varying rates
Cancellation, breakfast & payment differed per rate
B2C expectations
Users wanted a Booking.com-grade flow, not a spreadsheet
Higher drop-offs
Cognitive load → exits during room selection

The rooms page felt cluttered, created high cognitive load, and led to drop-offs during room selection.

I just want one Deluxe King, why are there nine of them?
Frequent booker, hotel L2 survey
02 · Design challenge

How might users quickly compare rooms without losing rate transparency?

The trap was easy: hide complexity, lose transparency. The win was harder, collapse the visible noise, surface what changes, and respect B2B reporting constraints at the same time.

Solution · Room ↔ Rate hierarchy
03 · Solution

Group the rooms. Expose the rates.

I redesigned the hotel room selection experience by grouping duplicate rooms across vendors, separating room-level and rate-level information, and introducing smart, hierarchical filters.

The result reduced visual clutter, made rate comparison faster, and aligned the experience with industry-standard booking patterns, while respecting B2B constraints around audit, policy, and reporting.

Before · 25-30 cards
After · 6 rooms

Five moves that shaped it.

Each move was small and unsexy on its own. Together they turned a wall of cards into a real shopping experience.

01

Room deduplication

  • Grouped rooms by name, bed type, and view across vendors
  • Displayed shared attributes once at the room level
  • Surfaced multiple rates under each room for easy comparison
  • Reduced visible room options from ~30 to ~6 per hotel
02

Clear room-level vs rate-level hierarchy

  • Room: name, bed type, images, view, prioritized amenities
  • Rate: price, cancellation, payment type, rate type, breakfast, loyalty
  • Helped users scan what stays constant vs what changes
03

Smart, hierarchical filters

  • Six room-level filters, loyalty, cancellation, payment, bed, breakfast, rate type
  • Supported multi-select combinations
  • Greyed out unavailable options instead of removing them
04

Priority-based information

  • Defined hierarchy for rate information based on NPS feedback
  • Replaced alphabetical amenities with priority-based ordering
  • Decision-critical info appeared first, not first-letter wins
Plate 02 · Vendor variability respected, user comfort restored 21 : 9 · Shop-floor view
04 · Designing within reality

The constraints that shaped it.

A clean dedupe model is easy on paper. In production, hotel inventory is messy, names mismatch, attributes vary, and audit trails matter. We made conscious trade-offs:

  • Chose not to show rate names due to inconsistent vendor data
  • Designed scalable patterns to absorb vendor and geo variability without rework
  • Kept full vendor lineage available in detail views for finance reporting
  • Greyed out unavailable filters instead of removing them, to preserve user mental models

The impact.

05 · Outcomes
−80%
Reduction in visible room options, from ~30 to ~6 per hotel.
Lower
Drop-offs on the Hotel L2 page through clearer hierarchy and grouping.
Reduced
Cognitive load & comparison effort, users scan, decide, book.
Higher
Booking confidence and conversion rates on hotel surfaces.

Field notes & learnings.

06 · Reflection
i.

User pain surfaces everywhere.

NPS, chats, emails, support tickets, sales calls. The signal isn't hidden, it's just spread thin. Triangulating sources was half the diagnosis.

ii.

Foundational UX problems need system-level fixes.

This wasn't a card style issue. It was a data + hierarchy + filter system issue. Visual polish on top of bad architecture only delays the rework.

iii.

Prioritisation is as critical as execution.

In complex products, what you don't show often matters more than what you do. Deciding which six things appear first was the design.

iv.

DQA is a design craft.

Pixel-level partnership with eng during DQA, naming edge cases, calling out drift early, meaningfully improved final quality.

07 · Takeaway

Reduce noise. Clarify choices. Respect how users think under complexity.

Designing for scale isn't about adding more features, it's about reducing noise, clarifying choices, and respecting how users think under complexity.

Next chapter · 04 / 04
Queen Bank, banking, quietly ↗
Open to opportunities

Like what you see?
Let's connect.

Status
Available
Q3 2026
Connect
LinkedIn ↗
Loc.
Bangalore,
India
© 2026 Jyoti Nagpal, All chapters reserved Designed in the dark · v.2026.01