Forecast-based financing is a financial mechanism that facilitates humanitarian actions prior to anticipated floods by triggering release of pre-allocated funds based on exceedance of flood forecast thresholds. This paper presents a novel model suitability matrix that embeds application-specific needs and contingencies at local level on a pilot project of forecast-based financing. The added value of this flexible framework is demonstrated on a set of hydrological and machine learning models. The model suitability matrix facilitates transparency and traceability of subjectivity in model evaluation. This paper advocates a stronger interface between model developers and end users for upscaling of forecast-based financing.