Analisis Prediksi Pembatalan Pesanan E-Commerce Menggunakan Algoritma Random Forest dan XGBoost untuk Mitigasi Kerugian Finansial

Ayla Zhafira; Fitri Yunita

doi:10.69693/ijmst.v4i3.12964

Authors

Ayla Zhafira Universitas Islam Indragiri
Fitri Yunita Universitas Islam Indragiri

DOI:

https://doi.org/10.69693/ijmst.v4i3.12964

Keywords:

Pembatalan Pesanan, E-Commerce, Random Forest, XGBoost, Kerugian Finansial

Abstract

Pembatalan pesanan (order cancellation) memicu inefisiensi operasional dan kerugian biaya tertanam (sunk cost) logistik pada industri e-commerce di Indonesia. Penelitian ini mengintegrasikan analisis biaya (cost analysis) ke dalam sistem prediksi pembatalan berbasis machine learning untuk menjembatani performa teknis dengan dampak finansial industri. Analisis komparatif dilakukan menggunakan algoritma Random Forest dan XGBoost yang dikombinasikan dengan teknik SMOTE pada dataset berisi 20.848 transaksi. Hasil pengujian menunjukkan XGBoost lebih unggul dengan akurasi global 86,63% dan recall 37%, serta berhasil mendeteksi secara akurat 209 transaksi batal pada data uji. Melalui simulasi parameter biaya, XGBoost terbukti menyelamatkan anggaran operasional sebesar Rp6.036.879, memberikan efisiensi moneter lebih tinggi senilai Rp1.117.800 dibandingkan Random Forest. Analisis feature importance mengonfirmasi bahwa opsi pengiriman, perkiraan ongkos kirim, dan lokasi geografis merupakan tiga faktor kunci pemicu pembatalan. Penerapan model sebagai sistem peringatan dini yang terintegrasi dengan strategi mitigasi proaktif, seperti verifikasi multikanal dan pembayaran pra-bayar, menawarkan solusi aplikatif untuk menekan kebocoran finansial sekaligus menjaga profitabilitas platform.

References

Aggarwal, K. K., Ahmed, S., & Malik, F. (2023). Optimal Pricing and Advertisement Policy for an Advance Order Booking Inventory System with Order Cancellation under Inflationary Condition. International Journal of Mathematical, Engineering and Management Sciences, 8(5), 869–895. https://doi.org/https://doi.org/10.33889/IJMEMS.2023.8.5.050

Andrianto, R. (2025). Bos BI Ungkap Transaksi Digital RI di E-Commerce Tembus Rp4.500 T. CNBC Indonesia. https://www.cnbcindonesia.com/market/20251030093858-17-680593/bos-bi-ungkap-transaksi-digital-ri-di-e-commerce-tembus-rp4500-t

Bhuria, R., Gupta, S., Kaur, U., Bharany, S., Rehman, A. U., Hussen, S., Tejani, G. G., & Jangir, P. (2025). Ensemble ‑ based customer churn prediction in banking : a voting classifier approach for improved client retention using demographic and behavioral data. Discover Sustainability, 6(28), 1–28.

Chawla, N. V, Bowyer, K. W., Hall, L. O., & Kegelmeyer, W. P. (2002). SMOTE : Synthetic Minority Over-sampling Technique. Journal of Artificial Intelligence Research, 16, 321–357. https://doi.org/https://doi.org/10.1613/jair.953

Jampur, F. X. A. P., & Christmantara, V. (2025). Analisis Dampak Metode Pembayaran dan Opsi Pengiriman terhadap Cancel Order Pada merchant Toko Mainan Edukasi Anak di Shopee. Jurnal Sistem Informasi (E-Journal), 17(1), 70–81. https://doi.org/https://doi.org/10.18495/jsi.v17i1.233

Jing, H., Zhang, H., & Wang, N. (2026). Deep Neural Mapping Unilateral Relaxed Support Vector Machines for Imbalanced Data. Discover Computing, 29(255), 1–21. https://doi.org/https://doi.org/10.1007/s10791-026-10148-3

Koten, F., Sudiarsa, I. W., Trisnawati, N. K., Pera, M. M. P., & Ndinin, M. A. (2026). Prediksi Pembatalan Pesanan E-commerce Indonesia Menggunakan Random Forest dan SMOTE. Jurnal Manajemen Akuntansi Dan Ilmu Ekonomi, 3(1), 8–14. https://doi.org/https://doi.org/10.70585/jumali.v3i1.192

Najafi-Ghobadi, Somayeh, Kalantary, R., & Mina. (2026). Optimizing Pricing in an Online Marketplace: The Role of Buy-Now, Pay-Later Option and Return Policies. Journal of Retailing and Consumer Services, 88. https://doi.org/https://doi.org/10.1016/j.jretconser.2025.104493.

Rafi, M., Faiz, M., Varshitha, K., Varsha, T. S., Lahari, K., Haque, M. A., Pagadala, P. K., & Dutta, S. R. (2024). Customer Churn Prediction employing Ensemble Learning.

Rodríguez-Garcia, M., Bas, A. O., Prado-Prado, J. C., & Lyons, A. (2023). Fulfillment Cost Analysis in Online Grocery Retailing: Store-Based vs. Warehouse-Based Strategies. International Journal of Production Management and Engineering, 11(2), 127–145. https://doi.org/https://doi.org/10.4995/ijpme.2023.18442

Roosevelt, M. W., Simarmata, J., Nofrisel, & Pahala, Y. (2024). Logistics Transformation in Indonesia for the E-commerce Market Industry. Malaysian Journal of Industrial Technology (MJIT), 8(3), 69–81.

Şahinbaş, K. (2022). Predicting Order Cancellations for E-Commerce Domain: A Proposed Model Based on Retailing Experience. Insan ve Toplum Bilimleri Arastirmalari Dergisi, 11(3), 1493–1514. https://doi.org/https://doi.org/10.15869/itobiad.1127578

Sarı, Z. S., Taşkapı, B., Dağdaş, G., & Akay, M. F. (2024). Development of Machine Learning-Based Sales Cancellation/Return Forecasting Models for the E-Commerce Industry. International Conference on Advances in Electrical- Electronics Engineering and Computer Science, 17–21. https://doi.org/https://doi.org/10.36287/setsci.19.4.022

Sarı, Z. S., Yurdagül, H. H., Taşkapı, B., Dağdaş, Gö., & Akay, M. F. (2025). Multi-Class Machine Learning Models for Predicting Order Cancellations and Returns in E-Commerce. 2025 9th International Symposium on Innovative Approaches in Smart Technologies (ISAS). https://doi.org/10.1109/ISAS66241.2025.11101817.

Tuba Irmak. (2025). Analysis of Order Cancellation Rates and Feature Weighting with Machine Learning: An E-Commerce Case Study. Intelligent and Fuzzy Systems (INFUS 2025), 1531, 763–770. https://doi.org/https://doi.org/10.1007/978-3-031-98304-7_82

Yonatan, A. Z. (2025). Tingkat Penetrasi Internet Indonesia Tembus 80% pada 2025. GoodStats. https://goodstats.id/article/tingkat-penetrasi-internet-indonesia-tembus-80-pada-2025-raped