SwapCircle - Peer-to-Peer Marketplace (Self-Initiated Venture)
A trust-first peer-to-peer secondhand marketplace for Uganda and East Africa. Built with Next.js 14, Supabase, and Prisma, with Mobile Money payments via an Aggregator, paid listings, identity verification, and automated moderation. All designed to replace informal social-media trading with a structured, monetized platform.
SwapCircle
(Peer-to-Peer Secondhand Marketplace)
Link: swapcircle.ug
Context & Market Gap
The secondhand goods market across Uganda and East Africa remains largely informal transactions driven by WhatsApp groups, Facebook pages, and word of mouth. Discovery is inconsistent, trust is thin, and there is no structured monetization layer.
Existing platforms that attempt to fill this space fail in predictable ways: no local payment integration, high operational dependency on manual processes, and architectures that don't survive beyond a single market.
SwapCircle was built to close that gap, not as a classifieds site, but as a scalable marketplace framework, designed from the ground up for East African market realities.
Objective
Build a digital-first, trust-anchored peer-to-peer marketplace capable of supporting regional expansion, with monetization and automation embedded from day one, and not retrofitted after growth.
Success criteria:
End-to-end self-service seller experience
Mobile Money as a first-class payment method
Revenue generation live from launch (paid listings, featured placements)
Automated moderation and listing workflows with minimal manual overhead
Architecture extensible to new cities and markets without structural rework
My Role
I initiated and owned the full product in terms of strategy, architecture, product management, and build execution. Across this project I operated as Product Owner, Solutions Architect, and Marketplace Strategist, with no external product team. All platform decisions, stack selection, feature prioritisation, UX flows, payment design, trust model, and go-to-market sequencing were made and executed by me.
Platform Strategy
The core design principle was automation-first, lean operations. The platform had to function with minimal human intervention, especially in early stages where operational capacity is limited.
Strategy execution focused on:
Structuring the seller onboarding flow around self-service and paid listing activation
Embedding monetisation into platform logic (not bolted on post-launch)
Building trust mechanisms - identity verification, human moderation, buyer reviews - as product features, not manual processes
Sequencing feature delivery in phases: core marketplace loops first, advanced discovery and analytics later
The go-to-market focus was supply-side first: make it easy and credible for sellers to list, then drive buyer demand through structured discovery and social amplification.
Technical Build
Stack: Next.js 14 · Supabase · Prisma · Tailwind CSS
Payments: Aggregator (Mobile Money primary, card supported)
Core modules built:
Module | What it does |
Seller Onboarding | Self-registration, profile setup, listing creation with Mobile Money payment activation |
Marketplace Engine | Listing management, category-based search and discovery, featured placement logic |
Payments | Aggregator integration; automated payment validation before listing activation |
Trust & Safety | Identity verification layer, human moderation queue, buyer reviews and ratings, reporting tools |
Operations | Workflow-driven moderation, automated listing state management |
The architecture is intentionally modular, each functional domain is decoupled to support phased feature expansion and future multi-market rollout without structural rework.
Design Decisions
Trust-first over volume-first. Rather than optimise for maximum listing volume at launch, the platform was designed to gate listings behind payment and verification. Paid listings filter out low-intent sellers. Human moderation catches what automation misses. This slows early supply growth but produces a cleaner, more credible marketplace.
Mobile Money as primary, not fallback. Card penetration remains limited across key user segments. Aggregator with mobile money as the default payment path, not a secondary option. This directly affects conversion for both listing fees and any future buyer-side transactions.
Custom build over off-the-shelf. Platform selection analysis evaluated Classima, HivePress, Kreezalid, Sharetribe, and Carousell clone scripts before committing to a custom stack. The decision was driven by the need for full control over data architecture, payment flows, and moderation logic, none of the evaluated platforms could support the trust-first model without significant compromise.
Designed Impact
Before | After |
Informal trading via social platforms | Structured, searchable marketplace |
No payment infrastructure | Mobile Money + card via Aggregator |
Zero seller accountability | Paid listings + identity verification + moderation |
No monetisation layer | Listing fees and featured placements live from launch |
Single-market, manual operations | Modular architecture ready for regional expansion |
Key Challenges
Balancing automation with trust. Automated listing approval creates risk, fraudulent or low-quality listings degrade the buyer experience and erode platform credibility quickly. The approach was to layer controls: payment activation as a first filter, automated content checks, and a human moderation queue for flagged items.
Avoiding premature complexity. Early-stage marketplace platforms frequently collapse under their own feature weight. The focus was held on core marketplace loops such as supply onboarding, demand discovery, trust reinforcement, and monetization, before expanding scope.
Local payment realities. Integration with an Aggregator required designing payment flows that handle Mobile Money confirmation latency, failed payment states, and multi-network coverage across Uganda and East Africa.
What I Would Do Differently
Conduct structured user testing on seller onboarding flows before finalising UX, specifically around listing payment friction
Introduce fraud detection signals earlier in the moderation pipeline
Validate listing fee pricing sensitivity across different seller segments and product categories sooner
Build in seller analytics from phase one — sellers who can see their listing performance stay on the platform longer
Link: swapcircle.ug