Adnova

Adnova is a smart, AI-driven marketing dashboard designed to enhance advertiser decision-making through deep campaign insights, user quality analytics, and real-time optimization suggestions. Powered by GPT-based prompt workflows and a modular UI system, the goal was to rethink how marketers interact with campaign data—making it intuitive, actionable, and scalable.

Year 2025
Collaborators Hamed Bahadori
Website hamedbahadori.com

Problem Statement

Advertisers and marketing teams often struggle with fragmented insights, unclear performance metrics, and generic dashboards that don’t adapt to their unique campaign needs. The challenge was to create a more intelligent, user-centric, and flexible dashboard that improves ROI, clarity, and confidence in decision-making.

Discovery & Research

Research Goals:

• Identify key user pain points in campaign management and optimization. • Benchmark current tools like Google Ads, Appsflyer, and Meta Ads Manager. • Explore how AI (especially GPT-based tools) can reduce workload and decision fatigue.

Research Methods:

• Competitive analysis • Heuristic review of existing dashboards • Expert interviews with performance marketers

Key Findings: • Users wanted fast ways to compare retention and ROAS across platforms. • AI suggestions needed to be trustworthy and context-aware. • UI needed to be clean, responsive, and prioritize clarity over decoration.

Design Approach

We structured the dashboard with a component-based, variant-driven design system, allowing effortless theme switching (dark/light) and scalable visual patterns across modules.

Key UI/UX Strategies:

  • Atomic design system with clear naming and autolayout structure in Figma
  • Accessible color contrast testing (WCAG AA)
  • Fully responsive layout for desktop, tablet, and mobile
  • Smart info hierarchy (AI alerts, campaign KPIs, and trend summaries at a glance)
  • Modular prompt builder to generate campaign insights without typing

Tools used: Figma, Stark, GPT-4, ColorBox, Notion

Structure & Navigation

Pages Designed:

  • Dashboard (KPI overview, AI alerts, user behavior graphs)
  • Campaigns (budget, retention, ROAS, and expert suggestions)
  • User Quality (segment-level insights and performance tracking)
  • Prompt Builder (custom GPT-based analysis generation)
  • AI Advisor (chat-based strategic suggestions)
  • Settings & Profile

Sidebar navigation was kept minimal and semantic, with clear iconography and path visibility. AI suggestions were embedded directly into context (not hidden away).

Visual Design System

  • Color Tokens: Neutral-based for clarity, with status and data highlights (green, red, purple)
  • Typography: Inter UI with consistent heading weights and hierarchy
  • Components: Reusable cards, data tables, toggle tabs, alert banners, modals
  • Variants: Light & dark themes powered by Figma variables
  • Microinteractions: Hover states, collapsible charts, tooltips, loading feedback

UX Flow & Interactions

  • Users land on a high-level dashboard with prioritized insights
  • AI Advisor answers performance questions via natural language
  • Prompt Builder allows dynamic analysis by selecting source, metric, time
  • Campaigns page reveals Smart Tips & Warnings contextually
  • User Quality shows retention graphs and segment breakdown

Animations and feedback were optimized for fast comprehension.

Impact & Outcome

  • Clearer campaign optimization workflow with reduced user cognitive load
  • Increased trust in AI outputs due to clear context and traceability
  • Ready-to-use UI kit for SaaS dashboards in marketing, fintech, or analytics

What I Learned

  • Variant-based systems future-proof multi-theme products
  • AI + UX only works if users trust the system's logic and context
  • Clear naming, tokenization, and Figma discipline pays off in handoff & iteration

Next Steps: • Add persona-based customizations for different advertiser types • Expand Prompt Builder templates based on marketer feedback • Publish the UI kit (Adnova UI) for community use and iteration