Originally Posted by CJJ
Pretty sure Kobo is a business built on harvesting personal data. And it seems to be getting worse. Google is no angel but then what corporation isn't mining information to get an edge? Good piece though. Thanks for info.
You're correct, they are harvesting data, and to an extent, most corporations are somewhat 'built on harvested data.' I'll reword that part so it's more accurate.
What I meant to say, was that Kobo's business is book sales. They do harvest this data and book buying patterns to optimize their book sales. But it is in service of that aim. Google on the other hand, makes 99% of their money from highly targeted advertising. Personal profiling & data mining is their business. Kobo isn't using the devices' GA metrics to personally profile users, but Google is
using them to target and profile individual users. (That's why Google provides GA)
Originally Posted by Lunchtray
It's not exactly counting button presses but sometimes yes. For instance if you have an option in your game that allows the player to change languages to any EFIGS and you notice it's never used that might indicate to not translate some games in the future or that you need to highlight it elsewhere in your marketing assets for other countries. There may be a menu button that you want players to use but they don't... what can be done to fix it?
Usually the analytics just grabs everything you can possibly do for a program and sends back the data and the company has to clean it up to understand how the customers are using the products and how to improve it.
I agree that detailed metrics can be informative, as in the examples you gave, but it's very possible to make a good product without them. Additionally, even if intensive tracking might provide greater insight, there's a limit to what is reasonable, no matter how 'anonymized
' or aggregated the statistics are. (I think this is something we both agree on, we just draw the line differently.)
I think Kobo adding an opt-in
toggle for usage statistics, with a statement of what they're tracking and the data's lifecycle, would allow them to still get good device data, while giving users a choice.
PS: I updated my post above so it describes your metric use better.