The Proper Use of Google Trends in Forecasting Models

It is widely known that Google Trends has become one of the most popular free tools used by forecasters both in academics and in the private and public sectors. There are many papers, from several different fields, concluding that Google Trends improve forecasts’ accuracy. However, what seems to be widely unknown, is that each sample of Google search data is different from the other, even if you set the same search term, data and location. This means that it is possible to find arbitrary conclusions merely by chance. This paper aims to show why and when it can become a problem and how to overcome this obstacle.

[1]  L. Gambacorta,et al.  Identifying regions at risk with Google Trends: the impact of Covid-19 on US labour markets , 2020 .

[2]  Matjaž Perc,et al.  Forecasting COVID-19 , 2020, Frontiers in Physics.

[3]  Futoshi Narita,et al.  In Search of Information: Use of Google Trends' Data to Narrow Information Gaps for Low-Income Developing Countries , 2018, SSRN Electronic Journal.

[4]  David E. Rapach,et al.  Now- and Backcasting Initial Claims with High-Dimensional Daily Internet Search-Volume Data , 2020 .

[5]  Rigoberto Pérez,et al.  Forecasting unemployment with internet search data: Does it help to improve predictions when job destruction is skyrocketing? , 2015 .

[6]  Thomas Mulder,et al.  Nowcasting New Zealand GDP Using Machine Learning Algorithms , 2018 .

[7]  Elisa Franco,et al.  The challenges of modeling and forecasting the spread of COVID-19 , 2020, Proceedings of the National Academy of Sciences.

[8]  Anna Simoni,et al.  When Are Google Data Useful to Nowcast GDP? An Approach via Pre-Selection and Shrinkage , 2019 .

[9]  Joni Heikkinen.  Nowcasting GDP growth using Google trends , 2019 .

[10]  Emilio Zagheni,et al.  Combining Social Media and Survey Data to Nowcast Migrant Stocks in the United States , 2020, Population Research and Policy Review.

[11]  Juri Marcucci,et al.  The Predictive Power of Google Searches in Forecasting Unemployment , 2012 .

[12]  Miguel de Carvalho,et al.  Real-Time Nowcasting the US Output Gap: Singular Spectrum Analysis at Work , 2017 .

[13]  C. Artola,et al.  Tracking the Future on the Web: Construction of Leading Indicators Using Internet Searches , 2012 .

[14]  Erik Christian Montes Schütte,et al.  In Search of a Job: Forecasting Employment Growth Using Google Trends , 2019, Journal of Business & Economic Statistics.

[15]  Andrey Fradkin,et al.  The Impact of Unemployment Insurance on Job Search: Evidence from Google Search Data , 2016, Review of Economics and Statistics.

[16]  N. Askitas,et al.  Google Econometrics and Unemployment Forecasting , 2009, SSRN Electronic Journal.

[17]  R. Tibshirani.  Regression Shrinkage and Selection via the Lasso , 1996 .

[18]  Marcelo C. Medeiros,et al.  Short-term Covid-19 forecast for latecomers , 2021, International Journal of Forecasting.

[19]  C. Anastassopoulou,et al.  Data-based analysis, modelling and forecasting of the COVID-19 outbreak , 2020, medRxiv.

[20]  Stefano Falorsi,et al.  Combining official and Google Trends data to forecast the Italian youth unemployment rate , 2017 .

[21]  A. Flahault,et al.  More Diseases Tracked by Using Google Trends , 2009, Emerging infectious diseases.

[22]  Nicolas Woloszko,et al.  Tracking activity in real time with Google Trends , 2020 .