Using Twitter to measure audience reception to a performance

This research was conducted by Deborah Agostino at the Politecnico di Milano, Italy.

Summary

This study looked at tweets associated with a 2015 run of performances of Aida at La Scala in Milan. It extracted and downloaded all the tweets in English and Italian that mentioned the name of the opera, the venue, the name of the conductor and the artists who were performing. The study picked out four indicators that could be used to analyse the audience’s reactions. The first was related to ‘content’ (the use of key words associated with the venue or performance), the second was ‘sentiment’ (whether people were being positive or negative), the third was the time at which tweets were posted, and the fourth was the Twitter users’ characteristics (if provided).

A process of data cleaning left 971 tweets to work with

82 per cent of the tweets in the study related to La Scala the venue but did not mention the opera. Seven per cent of tweets mentioned Aida and five per cent discussed the artists or conductor, one per cent mentioned the director. Half of the tweets were neutral, one third were positive, and the rest were negative. Less than 10 per cent of tweets were sent before the first performance, nearly 20 per cent after the final performance, while the majority were posted during the run.

There are benefits and drawbacks of using Twitter data to evaluate performances

Twitter was selected in this study because of its popularity and the ease of getting hold of the data. It can yield large data sets. The data does not have to be gathered in a survey or feedback form. It’s an organic listening exercise, rather than being directed by the interests of the venue or artist. People can be profiled as a result of the information they provide about themselves. One downside is that the analysis often requires advanced technical skills to complete, another is that the analaysis needs to be accompanied by proper appreciation of the contexts in which tweets are generated.