© 2022 Author(s).This study is mainly concerned with applying a fractional approach on the telecommunication sector in Turkey. We propose a model by using the fractional calculus and least squares method. This model aims to provide better performance for mathematical modelling of the discrete data. In the proposed model, firstly, the dataset is modelled with our previously developed fractional approach. Then, for better modelling, the same method is also utilized for error which is the difference between the dataset and the values found by the fractional approach model. By doing so, two layers' fractional approach gives better results than the Polynomial and previously developed Fractional Model. In this research, the fractional approach models are applied for mathematically modelling of the dataset. The dataset consists of the three different variables which are Gross Domestic Product (GDP) per capita of Turkey, average revenue per user (ARPU) of the mobile operators in Turkey, and special communication tax (CT) in Turkey. The dataset consists of the quarterly values of the variables with respect to years between 2011-2018. The correlation between the variables was examined. According to the results, we obtained the best modelling performance with the proposed model.