Product & Growth Strategy
Sentiment Analysis of Nigerian Fintech Apps
Millions of Nigerian fintech app reviews contain unstructured customer feedback that most institutions never systematically analyse. This project builds the methodological foundation to extract sentiment and thematic insight from that data at scale.
Year
2026
Location
London (remote research on Nigeria-based data)
Client
School of Finance and Managemnt, SOAS, University of London
Role
NLP Product Research Associate
Deliverables
LITERATURE SYNTHESIS, COMPARATIVE TOPIC-MODELLING
Duration
3 month
Approach
How It Came Together
Nigeria's fintech sector has scaled rapidly, but the customer-experience signal generated by that growth — millions of unstructured app-store reviews — remains largely unexploited by the institutions it concerns most. Working under Dr. Tolu Lawal, I led the research design for a text-analytics study examining customer sentiment and thematic patterns across Nigerian mobile banking and fintech applications. The engagement began with a comprehensive literature synthesis spanning mobile money adoption theory (TAM, UTAUT), regulatory and security determinants of trust, and precedent text-mining studies — including comparable analyses of over 37,000 Nigerian banking app reviews and 230,000+ reviews across payment platforms — to establish a defensible methodological baseline before any data collection began.
Rather than default to a single technique, the design phase evaluated three competing topic-modelling approaches — Latent Semantic Analysis, Latent Dirichlet Allocation, and Structural Topic Models — against precedent applications in adjacent review-heavy industries (hospitality, airlines, tourism) to select the methodology best suited to the sparsity and informality of Nigerian app-store language. This benchmarking-before-building approach mirrors the same principle applied in the SOAS AI-search engagement: methodological rigour up front reduces the risk of analysis that looks sophisticated but fails to generalise. The resulting framework positions the research to surface actionable sentiment and thematic drivers — service reliability, trust, responsiveness — that Nigerian fintech providers can act on directly, rather than a purely academic contribution.
Projects
