Resumen
The article discusses the modeling of artificial intelligence to address the issue of predicting future demand in the service sector (hairdressing, massages, exercise, doctors, physiotherapists, etc.). Future demand is examined on a weekly granularity and may exhibit fluctuations, particularly declines. The intention is to proactively respond to these fluctuations through marketing automation and attract additional customers who will compensate for the decline in demand. The result is a combined artificial intelligence that predicts future revenues for entrepreneurs using time-series analysis (revenues are derived from demand). Subsequently, a binary classifier is applied to identified demand fluctuations to determine numbers of future customers. Thanks to artificial intelligence, marketing automation is made possible. The classifier minimizes the numbers of false positives and false negatives while maximizing true positives. The goal is to maximize the utilization of labor and capital for entrepreneurs, as well as optimize their marketing expenses. Discount offers will be sent to only a limited number of customers where there is a higher probability of response. Online advertising will only be operational for the necessary duration, allowing entrepreneurs to achieve cost savings.