Daily Fantasy Sports Growth & Smart Recommendations
Diagnosing an engagement decline and designing personalized contest recommendations for an African DFS platform.
Squads is a Daily Fantasy Sports (DFS) platform where users draft real athletes into short-term contests for prize pools. This work covered three fronts: the strategic and regulatory landscape of DFS expansion, a data-driven approach to diagnosing a decline in engagement, and a full product specification for a personalized recommendation feature designed to lift conversion.
DFS sits in a regulatory grey area between skill gaming and gambling, demanding robust age/geolocation controls and responsible-gaming safeguards. On top of that, the platform faced a decline in engagement with no clear root cause — acquisition, retention, a technical payment failure, or a weakness in the contest offering itself? And growth depended on expanding beyond football into new sports without alienating a price-sensitive, trust-cautious user base.
How it came together
Research & Analysis
Mapped the DFS model, its key risks (regulatory compliance, responsible gaming) with concrete PM mitigations, and built a sport-expansion strategy prioritising basketball then cricket against demand, data availability, and regulatory factors.
Data & Metrics Diagnosis
Defined a prioritised metric investigation — login/session, browse-to-entry funnel, payment success, contest participation by type, and new vs. returning activity — to separate a retention problem from a product-offering problem.
Product Specification
Specified 'Smart Contest Recommendations': an ML-driven home-screen surface with goals, KPIs, user stories, acceptance criteria, a phased go-to-market, and edge-case handling.
What we shipped
- Smart Contest Recommendations — 3–5 personalized contests surfaced above the general list, updating in real time from user activity and live contest availability.
- Beginner-friendly onboarding for new users via educational tooltips and a popular-contest fallback when user data is sparse.
- Responsible-gaming controls: deposit/loss limits, cooling-off periods, behavioural monitoring, and self-exclusion tools.
- A phased rollout (20% beta → 50% → 100%) with A/B-tested recommendation algorithms and in-app feedback loops.
Results that mattered
target browse-to-entry conversion (8 weeks)
target time from login to first entry
target weekly sessions per user
target new-user 7-day retention
The core insight was sequencing: before building anything, instrument the funnel to learn whether the decline is acquisition, retention, or offering. The recommendation feature only earns its place once the data shows engaged users dropping off at the browse-to-entry step — otherwise it's a solution in search of a problem.