 Demand forecasting models for local supermarkets were developed and compared.  A dataset obtained from internet sales and mobile sales was used.  The best prediction result was obtained using Long and Short Term Memory Networks Figure A. 10-fold cross-validation results Purpose:The demand for a certain category of products is estimated by taking into account the e-commerce data (website and mobile application) of a local supermarket for the last two years and the factors affecting product sales (CPI and unemployment data). Theory and Methods: Twenty-four different methods of six different artificial intelligence algorithms (i.e. Deep Learning, Artificial Neural Networks, Gaussian Process Regression, Regression Tree, Support Vector Regression and Ensemble Learning) have been trained and tested for the demand forecasting model. Results:The obtained results were evaluated using correlation coefficient (R), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) criteria. The best result was obtained using Long Short-Term Memory Networks (RMSE= 0.0353; MAE= 0.0164; R = 0.9742). Conclusion:The results obtained can increase the success of e-retail sales by using the correct number of product orders, sales campaigns, and marketing strategies.