Title: Forecasting social media engagement: an ARIMA-based approach

Authors: Samer Abaddi

Addresses: IDentity Research, Amman, Jordan

Abstract: This study analyses social media engagement on X Company's Instagram and Facebook pages, forecasting the engagement rate (ER) using the autoregressive integrated moving average (ARIMA) model based on 2 years of data. The objectives were to understand social media behaviour, develop and evaluate an ARIMA model for ER forecasting, and develop a predictive social media dashboard. Findings revealed high predictability for both Instagram and Facebook ER data, though the high mean absolute percentage error (MAPE) and symmetric mean absolute percentage error (sMAPE) indicated room for improvement. Using business intelligence tools (Power BI), the study has developed an interactive dashboard for real-time insights and reliable forecasts, enhancing social media management for X Company. Continuous monitoring and adjustment of strategies based on actual ER values are recommended due to possible real-world changes.

Keywords: ARIMA forecasting; engagement rate; social media analytics; predictive dashboard; content type; Stata; Power BI.

DOI: 10.1504/IJASM.2026.150556

International Journal of Agile Systems and Management, 2026 Vol.19 No.1, pp.49 - 80

Received: 29 Jan 2024
Accepted: 18 Sep 2024
Published online: 17 Dec 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article