In 2021, the commerce sector in Brazil emerged as the primary source of job creation, responsible for around 26% of jobs in the country. However, the restrictions adopted to combat the COVID-19 pandemic affected the retail trade during 2020, resulting in a 1.4% reduction in sales volume compared to revenue levels seen in 2013. This paper analyzes and forecasts the monthly nominal revenue index for the retail sector of household appliances using the Holt-Winters, Sarima, Keras, Facebook's Prophet, and Random Forest models. The index presents a seasonal behavior that repeats every year, with  higher revenue values in the last quarter of each year, period marked by major sales promotions and festive occasions. The time series is model during the pre-pandemic period and during the pandemic period. The Holt-Winters model with multiplicative seasonal and additive trend components demonstrated the best performance using the MSE and MAPE criteria with a mean absolute percentage error of predictions of 3.95. Finally, it was observed that considering the index observed during the pandemic period in model training led to greater accuracy in predictions.