Improving visual consistency of admin panel which manages 12K+ devices — redesigning style settings for better branding control
My Role
Solo Product designer
Timeline
10 weeks
Team
Product owners, Support team
Platform
Web (Admin Panel)
Overview
AMS (Admin Management System) is a B2B platform managing 12,000+ restaurant and café devices worldwide.
Style Settings is one of the most critical modules in the product: it controls branding per location and is a required step during onboarding. Despite this, it consistently generated support requests and caused confusion for new clients — making what should be a self-serve flow dependent on manual support intervention.
Challenge
4 out of 6 new clients struggled with Style Settings during onboarding and required support assistance
4 out of 6 new clients struggled with Style Settings during onboarding and required support assistance. The root cause: the module had no logical grouping. Color palettes, media uploads, and zone configurations all lived on a single unstructured page — creating cognitive overload for users trying to set up branding for the first time.
Pain points from async support team survey
👩🏻💼 I have to upload the same 100+ images for every device separately, is there a way to reuse them?
🧑🏼💼 I don't understand how the image is uploaded, why it is cropped
🧑🏻💻 What does "Primary color" mean?
👨🏼💻 How can I crop the images correctly? I don't see any preview
👩🏻💻 I can't tell which styles are applied to which zone
Before: one chaotic screen
After: modular system with reusable assets
My role
Full design ownership
I initiated this project independently — identified the problem, structured the research, formulated hypotheses, and designed the concept — while running regular syncs with product managers and the support team to validate direction. Delivered in 10 weeks alongside other active projects.
Process
Research
With no direct user access, I focused on the people closest to the problem: the support team. I designed async written surveys to map recurring issues and pain patterns. I cross-referenced findings with product team knowledge about client onboarding.
This gave me a structured problem space to work from — without guessing.
Customers upload all their images and then accidentally exit without saving — and they have to start over from scratch
Anton · Lead Support Manager
New clients just pick a color, save it, and then call us because they have no idea where that color actually appears on the screen.
Veronika · Support Manager
They load a logo and it just doesn't show up on the device. No error, nothing — they don't know what went wrong
Yulia · Support Manager
Some clients don't even try to upload their media. They just send us their brand kit and ask our designer to set everything up on their behalf
Ilya · Support Manager
Hypotheses
If we separate palettes, media content, and preview into distinct tabs, then users will find the settings they need without getting lost, leading to faster module setup and fewer support requests related to navigation.
We believe that a reusable style library will help multi-location clients eliminate repetitive manual work. We will know this is working when we see clients applying saved presets instead of re-uploading assets for each new location.
If we add a live preview showing how branding changes appear on actual devices, then users will self-correct mistakes before saving, leading to fewer support requests about image cropping and color rendering
Design Decisions
The redesign introduced a modular tab structure — Colors, Media, Preview — each with a clear purpose and scope. Key decisions:
Tab navigation
Replaced the single-page dump with focused tabs. Each tab has one job, reducing cognitive load and making the module scannable
Reusable style presets
Clients can create named palettes and media libraries, then apply them across locations — eliminating repetitive uploads.
Live preview panel
A persistent preview shows how branding changes look on actual device screens before saving — removing the need for support-guided corrections.
Results
Qualitative outcomes:
Concept presented to C-level and used as pitch material for partner discussions — demonstrating the product’s redesign potential,
Provided the business with a scoped effort estimate for a full module redesign as a starting point for roadmap planning,
Established a design direction that could serve as the visual foundation for a broader AMS system update,
Reduced ambiguity for the product team: the concept made abstract UX problems concrete an decision-ready.