Forecasting crowdfunding platform revenues using ARIMA model

Abstract

Purpose – In recent years, crowdfunding platforms have become very popular as intermediaries between fundraisers and funders. However, various campaigns published on the platform might be of bad quality or fraudulent, so the crowdfunding platform must be very careful when choosing the right ones. Also, the proper selection depends on the profits of a crowdfunding platform. In most cases, crowdfunding platforms mainly earn money from transaction and administration fees from successful campaigns on their platforms. While it is very hard to select successful campaigns, it is possible to analyse already published campaigns and forecast future revenues of crowdfunding platforms. And based on this, to select new projects which might be successful too. The aim of this work is to develop a model to forecast the revenues of crowdfunding platforms. Research methodology – In this research, comparative and statistical analysis will be used, data structuring, modelling and forecasting, performed with the ARIMA model. Findings – Main findings of this research present the three most successful campaign categories from the Kickstarter platform that receives the highest funding. Fees were calculated from those three campaign categories, and revenues for the platform were forecasted using the ARIMA model. Research limitations – Main limitations are that there were used data from a very short period of time. For better results accuracy, a longer period is needed. Practical implications – this research might be of practical use since the forecasts show that the revenues will continue to grow from the successful campaign categories. Consequently, investors should continue to support technology, games and design categories the most. At the same time, crowdfunding platforms should give more attention to these categories when choosing new projects to launch on the platform.