Location-Based Services: Design with Caution
By Douglas K. van Duyne on August 11th, 2009Long the holy grail for smartphone developers and recipients of much fanfare and funds, location-based services (LBS) present a number of big design challenges when it comes to sharing location information. Some LBS applications provide useful utilities by reading the user’s current location and using that information to provide useful location-based information to the user, like Google Maps, the mapping and directions service; Yelp, yellow pages plus business ratings; Nike+, a log that keeps track of how far users run; GyPsii, a diary that stores text, location and images; and cool augmented reality applications like AcrossAir’s London Tube Finder.
However, some location services that share users’ location information on a network, including those built by venture-backed upstarts with funds to the tune of $20 million, actually share location information without users’ knowledge or understanding. These services offer answers to all-important questions like “which of my friends or colleagues are nearby?” and “Here’s a picture of my living room at 188 Chestnut Street, in San Francisco.” These applications often fail to address the interface and usage challenges as they tread on privacy, technology, or social network potholes.
Privacy Potholes
Privacy presents a challenge to LBS application design since people don’t always wish to disclose their location, and want to control to whom and under what conditions to disclose that information. Applications typically do a bad job of informing users, and of reminding them, when this information is being shared and with whom. Doing a search on Twitter for “location iPhone” will reveal that people are tweeting their location to everyone, and probably unwittingly. Other examples include TwitPic, which tweets your exact latitude and longitude along with pictures, and some Flickr apps save their location, too. That fact is often lost to users. It is somewhat of a stalker’s dream to Google or search on Twitter for somebody, find their current location or images of their children or other loved ones, and know their location. Think of a thief who sees pictures of the interior of your house on Flickr and knows where you live. Is it possible he could just go shopping online from Flickr and come rob your house when you Tweet that you are on vacation? Application designers need to better inform users, and let users control what private details they are providing and to whom.
Technology Potholes
Friend-finding apps are trying to connect people to their close network of friends. But these iPhone apps suffer from the limitation that iPhones do not broadcast location once the application is no longer running. This leaves one’s network in the dark for the vast majority of the time, since no one keeps an app open for very long. Android applications can run in the background, but iPhone applications do not. The consequences for the battery life on these devices, if they are always updating their location in the background, may become an issue, too. Until these technology hurdles are overcome, we won’t truly know how useful it might be to always know where our friends are right now, or how creepy it might be to feel watched. The killer friend-finding application is yet to be designed.
Social Network Potholes
The location-based social networking services are like unconnected islands that require circles of friends to download and use the exact same application. Until a central repository, possibly provided by Facebook, Google, or Apple, allows everyone to share information more seamlessly, these location-sharing LBS applications will remain islands with few inhabitants and resources. Getting one’s community to agree on which application/island to join, and getting them to use that application, is a big problem. I predict none of these systems will take off until there is a shared repository that all the applications can utilize.
Written with Mary-Anna Rae.
You make some interesting points. First off, I don’t think Facebook can be considered as an LBS as it does not incorporate any location information. Yes you can join a regional network but classing it the same as a true LBS such as Locassa, Brightkite or Latitude currently is incorrect; maybe they will change their user model at some point but for now it is a linear time-based social network.
You make a valid point about the iPhone and the single-task limitation when using LBSs but equally this aids in the privacy feature of such networks running on this particular device. I know such services will only be accessing my location when the application is running, hence eliminating some of the concerns you raise in your second paragraph.
We are very strong on privacy issues and are putting strict controls in place to put the choice for information disclosure firmly in the hands of our users. We provide the tools to share information but the privacy level that is employed should always be governed by the person adding the information, not the network owners.
I disagree that social networks need to have a common repository to share information before these services are truly useable, in fact all you need are comprehensive open APIs for each service that allow 3rd party developers to bring these networks together or for existing network owners to integrate other content feeds alongside their own. Many of the top social networks have such APIs and as such there are a few applications that bring these content feeds together.
There is going to be a shift in LBSs in the coming months, these networks have been too focused on a linear friends model and thereby limiting their services and target markets without thinking sensibly about the potential for geographic based services.