Return to Blog

Exploring the Challenges of Location Data Privacy

The internet is buzzing with the changes that iOS11 and Android Oreo bring to location data privacy. In iOS11, a prominent blue bar may appear when an app is using GPS to get a location fix, making it very easy to figure out if an app is doing higher fidelity location tracking in the background, without providing end-user value. In addition, iOS11 will also give the user more options and flexibility in location permissions. Instead of “Always On” or “Never,” users can also select “While Using App.” Android O takes a slightly different approach towards the same end, by limiting the number of background location requests to just a few per hour.

Both Apple and Google are sending strong signals to developers that they should be providing end users with commensurate functionality if they’re requesting location. And rightly so — location data can be incredibly sensitive, and the public is increasingly aware of who is tracking their location and how their data is being used and monetized.

As a developer, what are some things that you should think about to protect your users’ privacy and retain their trust? Here are some basic questions to ask:

  • Is location data integral to the core value of your app? Could you be asking for more than what you need to solve a problem? For example, you may not need a user’s exact location to suggest movies they may want to watch.
  • Is there a clear cut relationship between what you are asking for and what you are delivering? For example, people understand that route navigation apps need to be able to access location in the background in order to suggest routes when traveling, and not to lose track of the route when the user switches to their music app.
  • Do your users fully understand what data your app collects and why? Transparency is important.

If you’re still reading, chances are user location data is important to your app.

One approach to privacy is to simply store user data on the device and never transfer it to the company’s server1. The tradeoff of keeping data on-device makes it more difficult to access the data for bulk operations like model training, finding insights, and other development purposes. However, recent research from Google introduces a potential have-your-cake-and-eat-it-too methodology with their Federated Learning model. It “enables mobile phones to collaboratively learn a shared prediction model while keeping all the training data on device, decoupling the ability to do machine learning from the need to store the data in the cloud.”

This is just the tip of the iceberg and an opportunity to shape the landscape on how the developer community might leverage data to make systems smarter, without compromising individual user’s privacy.


  1. Factual’s location intelligence Engine mobile SDK follows an on-device approach. Learn more about it on our website.