A data-driven advertising dashboard that makes complex campaign data instantly actionable.
Overview
AiQEM Tech is an Ethiopian AI and blockchain company providing advertising analytics services to businesses. As their in-house UX Designer, I designed an end-to-end advertising analytics dashboard that gave marketing teams and their clients a single place to track, analyse, and act on campaign performance data.
The product surfaced five core data types (impressions, click-through rates, campaign spend, audience segments, and conversion funnels) across a modular dashboard. The challenge was presenting that much data without overwhelming people who needed to make fast, confident decisions.
I was the sole UX Designer on the project, working across the full design process from initial research through to the high-fidelity Figma handoff delivered to AiQEM's development team.
5
Core data modules
2
User types served
1
Design system built
↓
Complexity, not data
The Problem
AiQEM's campaign managers spent significant time each week manually assembling data from separate tools to build client reports. The dashboard needed to remove that work entirely while serving two very different users at once.
The internal team needed to move fast: scanning across campaigns, spotting anomalies, and adjusting targeting in real time. Clients needed confidence that their budget was working, without having to understand the data underneath.
“The hardest problem wasn't choosing the right chart type. It was deciding what not to show, and when.”
Understanding the Users
The dashboard served two distinct groups. Designing for both at once without fragmenting the experience was one of the core UX challenges.
Campaign managers
Needed to move fast: scanning across campaigns, spotting anomalies, and adjusting targeting or spend in real time.
View-only access
Not deep analytics users. They needed confidence that their budget was working and their ads were reaching the right people.
Design Process
Interviewed AiQEM's campaign managers about their daily workflow: how they moved between tools, what decisions they had to make quickly, and where the friction was. The key finding was how much time went each week into assembling data by hand. The dashboard needed to remove that.
Audited Google Ads, Meta Ads Manager, and HubSpot's analytics, looking at how each handled data density, filtering, and dual-user access. The best tools leaned on progressive disclosure and persistent global filters, two patterns I carried straight into the design.
Defined the module structure and navigation model before touching any UI. Key decision: a left-rail nav with five fixed modules, each containing its own filters and sub-views. A persistent global header with date range and campaign selectors applies context across all modules at once.
Lo-fi wireframes tested with AiQEM's internal team across 3 rounds. The most useful feedback: the first design surfaced too many chart types at once. I added a view-toggle pattern (table vs. chart vs. summary card) to every module as a result.
Built the full high-fidelity dashboard in Figma. Designed a complete component library covering charts, filter components, data tables, KPI cards, and modal patterns, with full developer handoff annotations.
Design Decisions
Data hierarchy
Every module opens with a summary card showing the single most important number: total impressions, overall CTR, total spend. Detail sits one click away rather than on screen by default. Someone could scan the whole dashboard in under ten seconds for a health check, then drill in where needed.
Filtering system
A global campaign selector and date-range picker live in the top navigation, and any filter set there applies to every module at once. That solved the orientation problem: people always know what they are looking at is consistent across views.
Chart language
I defined a consistent chart grammar: time-series data always uses area charts, breakdowns always use horizontal bars, funnels always use the same step-down shape. People learn the visual language once, and after that pattern recognition makes moving between modules quick.
Dual-user design
Rather than building separate interfaces for the internal team and clients, one dashboard adapts to permission level. Internal users see all campaigns, clients see only theirs, and the underlying UI is identical, which kept design and build simpler.
Outcome
The dashboard was delivered as a complete Figma handoff covering all five data modules with a fully documented design system, component library, and annotated specifications.
AiQEM needed to ship before a competitor reached the market, which meant making fast, well-reasoned calls instead of over-deliberating. Progressive disclosure and persistent filtering solved the data-density problem cleanly, and happened to be the most buildable option too.
It pulled together what used to take several separate tools into one coherent experience, and gave AiQEM something they could confidently demo to clients.
