We have in Humansensing argued that the new empirical source: sensor data, ecological momentary assessment and log data should be paid far higher interest in research.
Regarding the last empirical data source, log data they document ecological digital behaviour, you can analyse on a segment level and move back to individual level. The researcher can have the choice.
The other argument for the usage of log data is the fact that they are cheap. The disturbance of the user is minimal and in some cases the resources spent on preparing equipment for is small.
All these argument seems to be suitable for Apple excerpt for the last.
"Apple buys startup Topsy; gets rich Twitter data" says the news from Reuter(Dec 2, 201).
Apple buys a compagny, Topsy, solely says the story, to get their hands on all twitter data from the start of Twitter till now. Few compagnies have that access and it is valuable for Apple because they cannot track(that is news for me) their own users usage of their mobile Apps.
The point is the prize of the access to the twitter data:
200 mill dollars.