Value proposition operationalization in peer-to-peer platforms using machine learning

The purpose of this paper is to operationalize the value proposition in peer-to-peer platforms, by analyzing from all the variables which ones contribute the most for being an Airbnb Superhost. Authors use two different Machine Learning methods: Boruta for feature selection and SVM classification for prediction. More than 250 variables from 5136 listings were analyzed in the Canary Islands region. Results indicate that the Peer-to-Peer Platform Value proposition can be decomposed into three components: shared resources, value package and communications. Value proposition operationalization shows the possibilities and contribution of Machine Learning in the field of Tourism and Marketing. As practical implications for hosts, relevant variables help to have an understanding of the potential not addressed in their own value proposition. For Airbnb, relevant variables could be highlighted in search results or filters. For other companies, relevant variables of the value proposition can help to operationalize.

Automated summary

This paper aims to operationalize the value proposition in peer-to-peer platforms by analyzing which variables contribute the most towards being an Airbnb Superhost. Using methods like Boruta for feature selection and SVM classification, over 250 variables were analyzed in the Canary Islands region. Results suggest that the value proposition in peer-to-peer platforms can be broken down into shared resources, value package, and communications. The application of Machine Learning in the field of Tourism and Marketing shows great potential to help hosts understand untapped potential and highlight relevant variables in search results.