Ideajam
Kanban SaaS & White‑label Redesign.
Requests
Redesigned a Kanban SaaS platform to remove collaboration bottlenecks and build a white-label system adopted by enterprise clients without modification.
UI/UX Designer · Kanban Board · Design System
Restructured Kanban hierarchy and shipped a white-label design system that enterprise clients deployed without modification.
Steps to move a card from submission to review, post-redesign.
4 stages → 2 active
White-label system deployed by enterprise clients without any engineering rework.
From Scratch
Separated brand tokens from structural components for zero-friction white-labeling.
From legacy table rows
to structured visual hierarchy.
Every surface designed
for structured ideation.
The potential was there.
The navigation wasn't.
"It has the potential to be something like Jira — but I can't figure out how to navigate between my task and the other features. And I can't tell who's working on what."
Steps required to move a card from submission to review — each on a separate page with no persistent status indicator.
Pipeline FrictionShared design tokens in the original UI — colors, spacing, and radii were hand-coded per screen with no source of truth.
No Source of TruthDistinct visual patterns for buttons, inputs, and cards across the same product — each screen designed in isolation with no shared component language.
Pattern SilosCore visibility features absent at audit: contributor profile pages and per-idea member counts. Users couldn't see who created what or who was supporting it.
Missing VisibilityTotal scoring inputs per idea evaluation — 5 criteria × 6-point scale, presented sequentially with no progress indicator or guided completion flow.
Evaluator LoadMore navigation steps to switch between idea detail and the project board versus comparable Kanban tools. Context switching was the primary source of user drop-off.
Navigation DepthThree hypotheses.
Three confirmed decisions.
Reducing card progression from 4 explicit steps to 2 — with inline status indicators — would eliminate workflow confusion and cut time-to-review.
Collapsed 4-step flow to 2 explicit stages
Submission → In Review. Status surfaced inline on the card itself, no external doc needed.
Visual stage pipeline on Idea Detail page
Pre-Validation → EH Program → Approval Case → Incorporation — always visible at the top.
Persistent "Ideation Stages" panel on every idea
Progress tracker with team assignment and evaluation count visible without scrolling.
Hypothesis confirmed — users stopped losing track of card status after the inline pipeline launched.
Separating brand tokens from structural components would let enterprise clients deploy Ideajam under their own identity — zero engineering rework post-handoff.
Two-layer token system in Figma
Default Colourstyle layer (brand) fully decoupled from the component structure layer. Swap one, the other stays intact.
NBT white-label variant shipped as a parallel file
Enterprise clients received a second Figma handoff with their palette applied — no structural changes required.
Rounded corner + shadow system standardised
6px card, 4px normal, 40px radius buttons — documented and applied consistently across every component.
Hypothesis confirmed — enterprise clients deployed the white-label variant without a single engineering modification request.
Surfacing cumulative innovation scores inline — visible on the idea card — would reduce evaluator fatigue and increase scoring completion rates per idea.
5-dimension scoring panel
Innovation Degree, Feasibility, Viability, Strategic Fit, Sustainable Impact — each on a 1–6 scale.
Cumulative aggregate score surface
Single score (e.g. 4.3) visible on the idea detail and on the card in grid view.
Evaluation progress counter
"6 out of 8 evaluated this idea" — gives social proof and creates completion pressure.
Scoring completion improved in early enterprise cohorts. A forced-rank system would have been a stronger test — this is the main design regret (see Reflection).
Partially confirmed — completion improved but evaluator fatigue on 5×6 sequential scoring was not fully resolved.
Motion-first,
keyboard-accessible.
Main Design Elements
Core layout, card structure, and primary navigation patterns.
Create + Reply Thread
Idea creation flow, FAQ modal, and threaded discussion replies.
Mobile View
Responsive layout adapted for mobile-first users in the field.
The
Problem
The existing Kanban interface had no clear visual hierarchy between idea stages, causing users to lose track of card status and rely on external documentation. The product also lacked a white-label foundation, blocking enterprise clients who required brand-neutral deployments.
Process
& Strategy
Mapped user-reported friction points against session recordings to identify where cards stalled in the workflow. Designed a structured 4-stage visual flow and built a white-label design system from scratch, separating brand tokens from structural components so enterprise clients could apply their own identity without engineering rework.
Discovery & Audit
Mapped user-reported friction against 5 key screens. Catalogued 6+ inconsistent UI patterns and identified 2 missing visibility features — contributor profiles and per-idea member metrics.
Pipeline Architecture
Collapsed the 4-stage card flow to 2 explicit stages with inline status visible on every card — eliminating the need for external documentation to track where an idea was in review.
Token System & Handoff
Built a two-layer token architecture in Figma — brand palette fully decoupled from structural components. Delivered a white-label variant that enterprise clients deployed on Day 1 without a single engineering modification.
The
Solution
Delivered a restructured Kanban interface with explicit stage progression, visual status indicators, and a white-label component system built on separated brand tokens. Enterprise clients received a single Figma handoff they could deploy without modification.
What got
built.
White-label Support
Brand tokens fully decoupled from structural components. Enterprise clients swapped palette in Figma and deployed — no engineering changes.
Idea Preview on Create
Live preview pane visible while drafting an idea — users saw their card as it would appear in the Kanban grid before submitting.
Thread Reply
Nested reply threads on each idea post. Inline Reply action expands a sub-thread without leaving the idea card.
Global Channel Shoutouts
Broadcast promising ideas to the organisation-wide channel. Members outside the project get visibility on high-scoring concepts.
Detailed Progress View
Stage pipeline with evaluation count, team assignment list, and cumulative score — all visible on a single idea detail screen without scrolling.
Per-User Idea Profile
Each user profile shows ideas they pitched and ideas they are supporting — creating accountability and peer visibility into contribution.
Results.
Reduced the steps required to progress an idea from submission to review by 40%. The white-label design system was adopted by enterprise clients without any custom implementation requests to the engineering team post-launch.
Idea Progression −40% Steps
Enterprise White-label: 0 Custom Requests
0 Engineering Rework Post-Handoff
Design
System
User
Flow
Authentication and onboarding flow mapped before any visual design began — sign in, sign up, and credential recovery paths documented to surface where drop-off risk was highest.
What I'd do
differently.
30 scoring inputs
per idea.
5 criteria × 6-point scale, presented sequentially with no progress indicator. Evaluators completed the form but reported decision fatigue — especially on ideas with large team counts where everyone was expected to score independently.
A forced-rank model — drag 5 criteria into priority order rather than scoring each on a scale. One decision instead of thirty. Faster signal, no number fatigue, more consistent cross-evaluator results.
Evaluation panel
came too late.
The scoring UI was designed after the Kanban structure was already locked — it inherited the card pipeline's logic without its own usability pass. The 1–6 scale was never pressure-tested in isolation before handoff. The task-completion improvement was real, but the evaluator fatigue issue was always there and went unresolved.
Run a dedicated usability session on the scoring flow independently — before any visual design begins. Even a paper prototype with 5 evaluators would have surfaced the sequential fatigue early enough to change the model.