Case Study · 01

B2B Marketplace.

B2B Software Marketplace, End-to-end product design from scratch. First Runner-Up, AI Summit UAE 2025.

B2B SaaS Marketplace Head of Product Design Oct 2023 – Present Web · Desktop-first
Key outcomes
🏆
First Runner-Up, AI Summit UAE 2025, Best AI Product
10+
Major product features shipped from scratch
15+
Countries, enterprise buyers on the platform
6
AI-native features introduced, none existed before

This case study is protected.

The full case study contains confidential product details. Enter the password if you have it, or request access and I'll follow up personally.

or
Request sent. I'll follow up within a few hours.
Zoftware product showcase
10+ Features Shipped
6 AI Tools Built
15+ Countries
18 Months
01 The Problem

A product positioned around AI
with no AI features at all.

When I joined Zoftware in October 2023, the product was fundamentally broken. No AI-first features existed despite the company positioning around AI. The discovery flow had no logic, users could not reliably find relevant software. The homepage was not converting. No vendor onboarding system, no admin panel, no comparison tools.

The product lacked a design system, a component library, or any documented UX decisions. The brief was always informal: a problem statement, a feature idea, or a user complaint. Never a PRD. Never documentation.

I had to create structure, then create solutions within it, simultaneously.

"The platform kept growing. Every new feature was real value, but every new feature was also a new cognitive demand on users. My job was to keep expanding the product's capability while keeping the experience feeling simple."
Day-one problems
No AI features despite AI-first positioning
Discovery flow with no logic or relevance ranking
Homepage not converting visitors to users
No vendor onboarding system or admin panel
No design system or UX documentation anywhere
Concept overview
AI-guided discovery in a fragmented market

Buyers drowning in 15,000+ options. AI as the guide — not just a filter.

AI-guided discovery in a fragmented market
02 Role & Constraints

Sole designer.
Every workstream, all at once.

Role
Head of Product Design
Sole designer across all workstreams, product, marketing, vendor ecosystem, admin panel, and white-label.
Team
Cross-functional collaboration
Worked directly with development, product, marketing, and founding management teams daily.
Challenge
No precedent, no documentation
Continuously evolving product. Stakeholder requirements often in conflict with UX best practices.
Unique edge
Defend UX while shipping daily
Had to push back on stakeholder pressure while keeping the product moving forward, every single day.
02.5 — Research · Competitive Analysis

Mapping the competitive
landscape.

In February 2026, G2 acquired Capterra, GetApp, and Software Advice from Gartner for ~$110M. Four of the five incumbent review platforms are now a single entity controlling an estimated 55–58% of global software-review influence. The category just consolidated — which is precisely the opening for an AI-first, independent, decision-support challenger.

Swipe to compare →

Criteria G2 Capterra GetApp SoftwareSuggest SourceForge Zoftware ↗
Scale 2M+ reviews, 2,000+ categories, 100M+ users 100k+ products, ~900 categories 2M+ reviews, 37k+ products 50k+ vendors, 1,500+ categories Largest dev/open-source directory 15,000+ curated listings
Ownership 2026 Now owns 3 rivals Owned by G2 Owned by G2 Independent Independent Independent
Positioning Dominant, review-volume + AI-cited Broad SMB discovery Feature-fit discovery Emerging-market reach Developer / IT / open-source AI-first decision support
Core strength Highest AI citation density (2.6× rivals) Broadest database Fit Finder weighted ranking Free listings, India reach Developer trust Tool-led guided discovery
Key gap Enterprise-skewed, juggling 4 platforms Less independent post-acquisition Low traffic (#95k global) Lower global authority Narrow to dev/IT, dated UX New entrant — building authority
AI capability Citations strong; discovery still review-led Minimal native AI Minimal native AI Minimal Minimal Strategy Builder · Fit Score · RFQ Builder · Smart Compare · Zain Bot
Monetization Vendor plans from $2,999/yr PPC $2/click, $500/mo min PPC, syndicated Free + paid promotion Listings + promotion Marketplace + advisory
UX signature 6-tab listings, sliding scores Clean, accessible Save-to-board Listing-dense Functional / dated Guided, AI-surfaced, decision-first
Key insight
The incumbents are all review directories — they list and rate, then leave the buyer to decide. The market just consolidated under one owner, reducing information diversity. No major player leads with AI-guided decision support.
Where Zoftware wins
It doesn't compete on review volume (unwinnable). It competes on getting the buyer to the right answer — via AI tools that turn "here are 200 options" into "here is your fit, your strategy, your ROI." Independent, AI-native, and MENA-rooted at the exact moment the category became consolidated and US-centric.
03 Core Design Challenge

15,000+ products. Three user types.
One coherent experience.

The platform has 15,000+ listed software products, multiple AI tools, a vendor ecosystem, and three distinct user types, buyers, vendors, and system integrators. The hardest design decision was managing cognitive load across a feature-dense platform without fragmenting the user journey.

The solution: a layered information architecture where power features are progressively revealed, not front-loaded. First-time visitors see a clean discovery surface. Repeat users and advanced buyers unlock complexity through contextual triggers, not upfront menus.

User Type 01
Buyers
Enterprise procurement teams evaluating software across 15,000+ listed products. Need smart discovery, comparison, and fit scoring.
User Type 02
Vendors
Software companies listing and managing their products. Need onboarding flows, listing management, and analytics dashboards.
User Type 03
System Integrators
Technical consultants managing multi-vendor deployments. Need deep product specs, API context, and ecosystem mapping.
04 The Insight That Changed Direction

Enterprise buyers don't browse.
They interrogate.

After analysing user behaviour and competitor platforms (G2, Capterra, Gartner Digital Markets), one insight stood out: enterprise buyers arrive with a problem, not a category.

The original marketplace structure assumed browsing behaviour. I redesigned the discovery architecture around interrogation behaviour, leading with smart search, system fit scoring, and strategy generation rather than category navigation.

Old approach
❌ Browse Architecture
Category-based navigation
Passive discovery by browsing
No relevance signal to user
High exit rate at product pages
New approach
✅ Interrogation Architecture
Smart search leading with intent
AI-generated System Fit Score
Strategy generation from user's problem
Decision-ready information hierarchy
05 What Was Shipped

10+ features.
Every single one from scratch.

Core Platform
3 features
AI-Native Features
6 features
Ecosystem & Backend
3 features
White-label & Support
2 features
Core Platform
01
Full Homepage Redesign
3 major revisions shipped live. Each iteration driven by user feedback and conversion data.
02
Discovery Flow
Category pages, trending software, region-specific listings across 15,000+ products.
03
Smart Search
Query-based software discovery with filters, budget, company size, industry, deployment type.
AI-Native Features (all introduced by me, none existed before)
AI 01
Strategy Builder
Questionnaire-based implementation strategy generator. Users answer 8–12 questions and receive a custom software strategy.
AI 02
System Fit Score
AI-powered software matching with a percentage fit rating. Shows buyers exactly how well a product matches their needs.
AI 03
Tech RFQ Builder
Deep questionnaire that generates a detailed tech stack requirement report for enterprise procurement teams.
AI 04
Smart Comparison Tool
Side-by-side comparison across pricing, features, integrations, deployment, market size, and competitive advantage.
AI 05
Zain Bot
LLM-powered software advisory chatbot. Users describe their problem; Zain recommends, compares, and explains.
AI 06
Web-Based Telecaller
Browser-based real-time support caller. Enterprise buyers connect with a specialist without leaving the platform.
Ecosystem & Backend
07
Vendor Onboarding Flow
Complete, registration, listing creation, profile management, analytics dashboard.
08
Admin Panel
Full admin design, vendor management, content moderation, analytics, user management.
09
White-Label Product
Separate white-label product flows for enterprise clients who wanted a branded version of the platform.
06 Outcomes

First Runner-Up.
G2 founder. 15+ countries.

🏆
First Runner-Up, AI Summit UAE 2025, Best AI Product
15+
Countries with active enterprise buyers on the platform
18mo
Zero to full multi-feature platform, daily shipping cadence
0
AI-native features introduced, none existed before
"Praised by the founder of G2, the world's largest B2B software review platform. The platform's AI features were cited as genuinely innovative for the MENA enterprise market."
G2 Founder · External validation · 2025
🏆
AI Summit UAE 2025, First Runner-Up, Best AI Product (all UAE AI companies)
G2
Praised by the founder of G2, the world's largest B2B software review platform
15+
Countries with active enterprise buyers on the platform
18mo
Zero to multi-feature, multi-user-type platform, daily shipping cadence maintained throughout
07 AI in This Project

AI made me faster.
The judgment stayed mine.

AI Handled
Research & Drafts
Domain research (procurement, enterprise workflows)
UX copy first drafts
PRD and documentation drafting
Stakeholder presentation structure
Human Reviewed
All Final Decisions
All design direction and UX logic
All stakeholder communication
Feature prioritisation and sequencing
Every shipped screen reviewed by me
100% Human
Craft & Taste
Visual design and visual craft
Design system decisions
Ethics and accessibility calls
All user-facing product taste
AI Tool Used for Human override
Claude Research on unfamiliar domains, UX copy, PRD drafting, design documentation All final decisions, all UX logic, all design direction
Figma AI Component variants, rapid iteration Visual judgment, final selection, quality bar
Generic AI Stakeholder presentations, data visualisation, sheet creation Strategic framing, narrative, business context
"AI helped me research complex domains I'd never worked in before, like enterprise procurement workflows and cost automation modelling. But every decision about how to present that complexity to a user, that was never AI's job."
Next case study
Eazy6
Real-Money Sports Contest Platform · Consumer · iOS + Android · USA
View case study