The increasing importance of services trade in the global economy contrasts sharply with the lack of timely data to monitor current trends. This information gap poses challenges for policy-makers during times of economic volatility, as has become evident during the COVID-19 pandemic. This paper presents a nowcasting approach aimed at providing insights on current developments in aggregate services trade, as measured by the monthly balance of payments. The approach combines regressions and supervised machine learning to select variables. The models incorporate hard, soft, financial and uncertainty indicators as well as data on Google search requests.