FanRush
Creator-led, performance-based marketing — an AI acquisition engine wired to a 5,000-creator network across 35+ countries.
FanRush is a performance-marketing platform that combines a curated network of 5,000+ creators with an AI-powered customer-acquisition engine. Brands set a CPA target and FanRush orchestrates organic creator content + paid amplification to hit it — with real-time routing, optimization and reporting. We deliver the engineering, AI surface and operator tooling behind the platform.
- AI acquisition engine
- Creator onboarding + payouts
- Brand campaign console
- Performance attribution + analytics
- Real-time bidding + paid amplification
- Scale
- 5,000+ creators · 850M+ audience
- Reach
- 35+ countries
- Customers acquired
- 1M+ per month
- Avg conversion
- 4%
- Pricing model
- CPA-based
Highlights from the
engagement.
- AI engine that routes campaigns to the highest-converting creator + channel mix in real time
- 5,000+ vetted creators across 35+ countries with combined 850M+ reach
- CPA-based pricing — brands set a target, FanRush hits or beats it
- Live attribution dashboards mapping conversions to creator, post and channel
- Brand console with offer scheduling, creative briefs and budget governance
- Creator payouts via Stripe Connect with multi-currency support
How we
delivered.
Brief, architecture decisions, what shipped and where it stands now — in our own words.
The brief
FanRush's leadership came out of search, traffic distribution and the creator-economy operator world — three decades of combined experience, much of it at publicly traded companies. They had a sharp thesis: most affiliate networks give brands access to offers and stop there; the next-generation play is to combine a curated creator network with an AI-powered acquisition engine that optimises across organic and paid channels in real time. They asked us to build that engine.
The clients we're acquiring for include some of the biggest names in commerce — SHEIN, Adidas, Zappos, Marks & Spencer, PrettyLittleThing, Macy's, Sephora — which means the platform has to behave like a serious mar-tech tool from day one. No demos. No 'we'll fix it in production'. CPA-grade attribution, live spend governance, payout-grade reporting.
Architecture & engineering choices
The acquisition engine is a hybrid Python + Node service. Python handles the AI: a bandit-style optimiser that learns which creator × creative × channel mix performs against a given CPA target, with a vector store of historical performance grounding the cold-start. Node handles the high-throughput attribution and webhook ingestion from brand checkout systems. PostgreSQL is the system of record for offers, conversions and payouts; a separate event store carries the live optimisation signal.
Stripe Connect powers creator payouts across 35+ countries with multi-currency support — we settled on Connect specifically because the network spans markets where domestic-only rails would have left half the creators in the cold.
What we shipped
- AI acquisition engine routing campaigns to the highest-converting creator + channel mix in real time
- Brand console with offer scheduling, creative briefs, budget governance and live CPA tracking
- Creator-side console with campaign discovery, content templates and earnings reporting
- Conversion attribution layer ingesting webhooks from brand checkout systems
- Stripe Connect payouts in 35+ countries with multi-currency settlement
- Performance dashboards mapping every dollar of conversion back to creator, post and channel
Where it stands now
FanRush is in production with 5,000+ vetted creators, a combined audience of 850M+, and 1M+ customers acquired monthly at a 4% average conversion rate — numbers that put it squarely in the serious performance-marketing tier. We continue to operate the engineering and AI surface, and the engagement has expanded twice.
PlaySQOR
Daily fantasy and real-money sports wagering, engineered for compliance and scale.