Protocol
The protocol adopted is quite similar to the previous one, unlike here we were interested in the most displayed images on Google. So, thanks to “isearchfrom.com”, we were able to navigate Google Images simulating to be in our 26 countries and using as query “Uber + country’s name in in the local language/alphabet”. Considering that an average Google Images page contains more than 15 pictures and less than 25, we stated to download the top 20 images displayed for each country, getting a final database of 520 images. We kept them in the same order as they appeared on Images. The images were downloaded thanks to the tool Down them all.
Finally, images were tagged as representing protest, police, Uber/taxi, only taxi, infographics, drivers, people and cars, only cars, the app, advertising or other stuff. All the images were also grouped according to the level of protest: promotional (no protest), low, medium or high level.