Centralized Design Library UX Case Study · 2026
Design System · 5 Products

One library. Five brands. One source of truth.

A single, scalable design library that powers five vehicle marketplaces. One component feels native to every brand, buyer and vocabulary, all from one centralized source.

The five verticals
Tractor Junction
Truck Junction
Bike Junction
Car Junction
Infra Junction
Role
Sole designer
Timeline
3 weeks
Tool
Figma
Products
5 verticals
Scale
40+ modules*
00 · Overview

Five marketplaces, one company, zero shared parts.

The Junction family runs five vehicle marketplaces under one roof. Each has its own audience, business goals and buying journey. But every product page is built from the same parts: a listing card, a spec table, an EMI tool, a review block.

Those parts had been built separately for each product. This project centralized them into one design library, designed solo, end to end, in three weeks.

Tractor Junction
Rural buyers. HP & CC. Research-heavy.
Truck Junction
Fleet owners. Payload & GVW.
Bike Junction
Riders. cc & kmpl. Quick decisions.
Car Junction
Urban families. Offers & EMI led.
Infra Junction
Contractors. Construction equipment.
01 · The problem

The same parts, five different ways.

01
Different buyers, one toolbox

Farmers, families, fleet owners, riders and contractors all shop through the same underlying components.

02
Different vocabulary

HP & CC, cc & kmpl, payload, bucket capacity. Same card, different words.

03
No shared foundation

Cards, buttons and grids were rebuilt per product. Inconsistency by default, and maintenance multiplied.

04
Slow to ship

Every new screen was built from zero, and a fix in one product never reached the others.

02 · Goals & guardrails

One library that stays consistent and flexible at once.

Consistency

Identical structure, spacing and behaviour on every product.

Flexibility

Room for each brand's colour, language and CTA to stay its own.

Speed

New screens assembled from ready parts, not built from zero.

Scalability

A new product can join by supplying tokens, not screens.

03 · The idea

Build the card once. Theme it per brand.

The insight that made one library survive five brands: separate what must stay the same from what is allowed to differ.

Fixed: the skeleton
Layout & grid Media slot Spec grid Price block CTA stack Behaviour
Themed: the tokens
Accent colour Spec vocabulary CTA verb Badge
04 · Process

Three weeks, three moves.

Move 01
Audit & tokenize

Collected every existing screen across the six products, mapped the overlap, and set the shared tokens (colour, spacing, radius, type) as Figma variables.

Move 02
Build the core

Built the container grid, the button primitive and the listing card as slot-based components, then themed each to every brand to stress-test the model.

Move 03
Scale to 40+ modules

Extended the system across the full product page: chrome, detail modules, the rail and discovery. Every module inherits the same tokens.

05 · Colour & buttons

One button primitive, five brand themes.

One structure: primary, outlined and linking, each with a hover state. Only the colour token changes between products.

Tractor Junction
Tractor Junction button set
Truck Junction
Truck Junction button set
Bike Junction
Bike Junction button set
Car Junction
Car Junction button set
Infra Junction
Infra Junction button set
06 · Layout grid

One container. One content and rail split.

Every product page sits on the same width. Content and rail keep the same split across all six brands.

Container
1280px
Content
898px
Rail
360px
07 · Impact

One source of truth. Five products, faster.

3wks
blank file to a shipping library
5
products on one system
40+*
reusable modules
~60%*
faster to build a new screen
* replace with your measured numbers
08 · Learnings

The hard part wasn't the components. It was deciding what six brands were allowed to disagree about.

Constraints scale

Locking the skeleton is what let six brands move without drifting apart.

Tokens over screens

Theming by variable meant a brand change was a token swap, not a rebuild.

Design for the next

Built so a future vertical can join by supplying tokens. The system is ready before the product is.