Researchers use tweets to predict iPhone sales figures
How may new iPhones are sold at launch? If your guess is 'a lot', then it is a lot less exact than a new method co-developed by Ravi Vatrapu, Professor of human computer interaction and Director of the Computational Social Science Library at CBS. This method is in fact able to predict the sales figures of the next quarter with a margin of 5 per cent plus or minus.
Together with students Niels Buus Lassen and René Madsen, Ravi Vatrapu has created a method on the basis of existing research, which is able give an unprecedentedly precise estimate of the iPhone sales figures. Apple's mobile phone has been used as an example, because it is one of the most hyped products globally.
20 days
Vatrapu's method adds up the total number of tweets discussing the product. The researchers are then able to see a connection between the number of tweets 20 days before the next quarter begins and the actual sales figures of the quarter.
- We say that a tweet about an iPhone shows that a user is attentive to iPhones. We don't know if the user is an existing customer, a potential customer or just a robot sending automatic messages about iPhones, but it is not important. What's important is that the user is attentive to the product, and that this attention always will be present in one of four stages, says Ravi Vatrapu to Videnskab.dk.
Vatrapu and his team will continue to improve the method.