Project Scope
- 1 Project Manager
- 1 Designer (me)
- A team of developers
01Problem Statement
Fragmented dashboards, confused sellers
Commerce and Venture home served different purposes but were experienced by the same sellers — making daily operations unnecessarily hard.
Sellers struggled because two dashboards co-existed with no unified logic for data, navigation, or permissions — making daily operations unnecessarily hard.
Core Issues
Inconsistent availability of data and features across different dashboards.
A centralized, integrated filter is needed to better filter and view charts.
UX Focus Areas
Dashboard with varying levels of access, navigation elements & filters, data representation for omni-channel scenarios:
02Project Goals
Two things we set out to solve
Based on the core issues, we aligned on two concrete goals that would define success for this project.
01
🔐
Establish role-based permissions
Define clear access tiers for different seller roles — from LLC-level owners managing multiple brands, to store managers scoped to a single location — and reflect these in the UI.
02
🎛️
Design a scalable filter system
Replace the existing single-select filter with a model that supports multi-location and multi-channel comparison — and can accommodate future omni-channel scenarios like BOPIS, BORIS, and ROPIS.
03Personas
Who we're designing for
Four key roles in the Yum Brands LLC hierarchy — from LLC owner to individual store manager — each with distinct access needs.
Yuri
Owns Yum Brands LLC for Taco Bell, KFC
Tobia
Managing Partner for Taco Bell — not Yum Brands LLC
Katrina
Managing Partner for Pizza Hut — not Yum Brands LLC
Silas
Store Manager at Pizza Hut @ Santa Clara
UX Model of Store Management
| Role | Allowed ✅ | Not Allowed ❌ |
|---|---|---|
| Brands Overall Ownere.g. Yuri | Access all brand data · View aggregated reports · Manage all products & billing | Full access (no restrictions) |
| Sub Brand Ownere.g. Tobias, Katrina | Access specific brand data · View brand-wide reports · Manage brand products & billing | Cannot access other brands |
| Store Managere.g. Silas | Access assigned store data · View store metrics · Manage daily store ops | No other stores' data · No billing/products · No brand/LLC settings |
Example
From permission model to screen
The model told us
Owner sees everything
Yuri needs a cross-brand home — aggregated metrics across Taco Bell, KFC, Pizza Hut in one view, with the ability to drill into any brand.
Manager sees their brand only
Tobias and Katrina land directly in their brand's dashboard — no cross-brand data, no access to billing or LLC-level settings.
Store manager sees one location
Silas only sees data for his assigned store. No other stores, no brand-level settings, no billing.
So we designed
A multi-business home for owners
Brand switcher + aggregated dashboard with cross-brand metrics. Yuri can navigate between Taco Bell, KFC and Pizza Hut without leaving the home.
A brand-scoped home for managers
Tobias lands on Taco Bell's dashboard only. The navigation and data surface is identical in structure — but scoped strictly to their brand.
Same IA, different data scope
All roles share the same navigation structure and component library — permissions control the data, not the layout.
03Low-Fidelity Model
Wireframes by role
Each role gets a scoped view of the same underlying dashboard — the permission model directly determines what's visible and accessible.
Owner — Yuri
Owns Yum Brands LLC for Taco Bell, KFC
Welcome back, Yuri.
Yum Brands
Taco Bell
Pizza Hut
KFC
All Businesses
Taco BellYesterday
KFC1 week ago
Pizza HutYesterday
Yum Brands LLC
Sales$1,500
Orders109
Customers88
Payouts$7,500.74
Owner view: Multi-business switcher at top + aggregated cross-brand dashboard. Payouts and Action Items visible across all brands.
Manager — Tobias, Katrina, Silas
Brand or store-scoped access only
Taco Bell
Taco Bell
Sales$4,000
Orders41
Customers72
Payouts$5,500.63
Manager view: Same dashboard structure as Owner — but scoped strictly to their brand. No cross-brand navigation, no LLC-level settings.
Low-Fidelity Model
Owner
Yuri
Owns Yum Brands LLC for Taco Bell, KFC
Manager
Tobias
Managing Partner for Taco Bell -- not Yum Brands LLC.
Katrina
Managing Partner for Pizza Hut -- not Yum Brands LLC.
Silas
Store Manager at Pizza Hut @ Santa Clara.
A new problem emerged
The next design challenge
The dashboard works — but
how do sellers slice their data?
Once we established who sees what, a deeper question surfaced: within their permitted scope, how do sellers filter and compare data across locations and channels to make meaningful business decisions?
Multiple Locations
Compare performance across locations to quickly identify underperformance and opportunities.
Multiple Channels
Sales span multiple channels—understanding which drive revenue helps prioritize where to focus next.
The filter becomes critical
Analyzing data across locations and channels requires more than simple filters—thoughtful filtering is key.
Why it matters
The filter directly affects how sellers read their charts and, in turn, the quality of the business decisions they make. A poor filter = misleading data = wrong decisions.
What we needed to answer
How should sellers select, combine, and compare across locations and channels — without cognitive overload? We ran targeted research and tested three filter proposals to find out.
Research & filter exploration below
About Filter
Proposal
1.
All Sales Channels
All Locations
Pros
- Allow location selection for clearer categorization.
Cons
- Not able to compare multi select in one filter.
2.
All Sales Channels
All Locations
Pros
- Allow multiple items to be selected in one filter for more flexible viewing.
Cons
- “All Sals Channels” and “All Locations” can be confusing.
3.
Pros
- Clear for users to understand.
Cons
- Takes up more space.
- Feels heavy for users who just want quick switching.
04Primary Research
Questionnaire and Interview
Two concept tests with 10 SMB participants to validate dashboard comprehension and understand how sellers mentally model “sales” and “sales channels”.
Key Questions
- How do we define different terms such as “sales,” “sales channels,” etc.?
- How do users comprehend the dashboard? Do they find it easy to use?
Method & Participants
10 SMB Owners
Test 1: Multi-location/multi-channel business-level dashboard with metrics, data viz. and filtering.
Test 2: Multi-business dashboard with cross-brand metrics and data viz.
6Retail & Services — multi-location & multi-business
4Construction, Retail, Pharma — multi & single business
10Total participants
2Key questions answered
Key Insights
What We Learned
Two core research questions shaped the tests: how do users define “sales” and “sales channels”, and how do they comprehend the dashboard?
Majority equate sale = payment. 7/10 said $25 should be reported as a sale as soon as the customer pays online.
“Sales channel” is confusing. Most users defaulted to “both in-person and online” rather than identifying a single channel origin.
Unpaid order ≠ sale. All 10 participants agreed an order placed without payment is not a sale — order source ≠ sales channel.
Current filter is unsuitable. Single-select dropdown doesn't support multi-location comparison or complex filtering scenarios.
05Prototype Test → Insights → Decision
Test — Multi-channel Dashboard
We built and tested a prototype to surface exactly where the filter broke down — then used those findings to evaluate three proposals before landing on a final design.
Step 1 — What we tested
Multi-channel Dashboard Prototype
Features
- 1.Separate locations & sales channels
- 2.Filtering of data based on locations and sales channels
- 3.Data viz. and metrics to support filtering and comparison
Test Tasks
1Comprehension of data
2Single location data
3Online sales for a location
4Location comparison
Step 2 — What testing revealed
⚠️
Current filter breaks down
The single-select dropdown forces users to switch repeatedly — making multi-location or multi-channel comparisons nearly impossible.
💡
Chart click feels more natural
Users intuitively tried clicking the chart to filter — suggesting the filter interaction should feel closer to the data, not separate from it.
🔧
Configuration over instant-switch
Power users preferred a “set up → apply → review” workflow rather than live dropdown changes that instantly altered charts.
Taking the best of each proposal → combining into one unified panel
Step 4 — Final Decision
Why the panel-based filter won
All filters visible in one panel → easier overview. Works more like a configuration, convenient for users who don’t need to switch often.
Easily accommodates future filters (date range, store type) and supports responsive design and accessibility.
Supports a “set → apply → review” workflow, aligning better with how users perform multi-condition data analysis than instant dropdown changes.
Filters
06Final Outcome
A unified Commerce & Venture Home
The final design resolved fragmentation, established clear permission tiers, and introduced a filter model validated through multiple rounds of testing.
Filters
Did we hit our goals?
Establish role-based permission pages
Three distinct access tiers — Brands Overall Owner, Sub Brand Owner, and Store Manager — are fully defined with explicit allowed and restricted actions, and surfaced clearly in the UI model.
AchievedDesign a scalable filter system
The panel-based filter replaced single-select dropdowns. Supports multi-location and multi-channel selection simultaneously, with architecture extensible to date ranges, store types, and future omni-channel scenarios.
Achieved🔐
Permission-aware dashboard
Three clear access tiers — Brands Overall Owner, Sub Brand Owner, Store Manager each with defined allowed and restricted actions.
🎛️
Panel-based filter system
A configuration-style filter panel replacing single-select dropdowns. Supports multi-location comparison and future extensibility (date range, store type).
📊
Unified data model
Resolved ambiguity around “sale” and “sales channel” terminology with consistent definitions aligned to how SMB owners actually think about their business.