BetaLabs

Reflections on Nokia Friend View

Written by James Reilly, Friend View Project Manager

Since the Friend View service was shut down in September, we’ve received a number of questions from avid users and onlookers who wanted to hear a bit more about what came out of this research project.  To give back to those that contributed, we did our best to gather some high-level observations and learnings that we could share with you here.

(For those that weren’t involved, Friend View was a social location and micro-blogging service that helps you keep up with your close friends. This research project from Nokia Research Center was published as an Experimental Beta on Nokia Beta Labs from 5.11.2008 - 2.9.2009.)

A few statistics

Over 11 months, Friend View grew to 30,000 registered user accounts, and about 2000 active sticky users. (An ‘active sticky user’ here approximately means a user account older than 30 days that was also used during the previous 30 days.) Because Friend View was an early phase Experimental research project targeted at close-knit social groups to share status and location, we expected that uptake and usage would be more limited than for other types of Beta’s.

Consistent with Beta Labs users, Friend View’s users were mostly male,technology leaders who have the latest phones and are willing to try new applications. Ideally the project would have attracted a more equal gender distribution and a wider usage distribution across different types of mass market consumer segments to enrich the possibilities of using the service with friends and family who are not necessarily early adopters.

A few observations

Here is a short summary of some observations that may be of interest:

• The social aspects of Friend View were used more frequently than location sharing, and seemed more appealing to users.
• Location and maps did however seem to add a nice element of ‘fun’ on top of the microblogging socialization.
• Friend View was primarily used on the go from the mobile phone, and not from its desktop browser web site.
• Users typically wrote short status messages most often. This may be because they generally used it ‘on the go’ (i.e. short pertinent messages, written while mobile).
• The idea of location sharing seemed to generally be polarizing, meaning users either found the idea appealing (at least for certain uses and situations) or they just did not.
• For those who had interest to try and use a service like Friend View, there was a stronger preference to use it with close friends and family than with their wider social circles or publicly.
• However, there were also requests to add more public features. Friend View could only be used within closed sets of users (sets of bidirectional friend pairs only), and had no “public” sharing features or channels.
• Some users did not always understand the difference between a very precise (GPS) and less precise (e.g. network-based (cell id) location) location, and why the precision changes when they go indoors and GPS signal is lost. That would have been an interesting UI and usability topic to have studied further in the Friend View client.
• Friend View had S60 3.0, 3.1, 3.2 and 5.0 Touch client versions available; the S60 5.0 Touch version was the most popular downloadable client.

Common feedback requests

The most commonly requested improvements were to:

• integrate with Nokia on-device and Ovi services like Maps, Contacts, etc.,
• integrate with other social networking services like Facebook and Twitter,
• make finding friends and friend invitations easier,
• support more mobile devices and platforms, and
• even better power management options/features.

There were also requests to support different sub-groups of contacts for sharing and privacy settings for those (i.e. better support for different friend circles), and in addition there were some requests also for more public features.

We often got questions about battery life. It’s an interesting topic continually improved since the start of the project. Some users wanted to be able to share location both ‘even more’ precisely and more frequently. Others were much more concerned about saving battery life, or perhaps for sharing location as a ‘higher level’ semantic location (e.g. City level). Still others may want to actively share their location throughout an entire moving session like their daily commute, but others only when they start and end their commute but not in between. This type of balancing act between what level of information to share, when and how often, for different types of users, activities and activity profiles was one interesting and challenging area for the project to work on.

The biggest challenge for many of Friend View’s Beta Labs users, however, was how to grow their social graph to be large enough to use it with their real everyday friends and family. This is reflected in user requests to support even more devices and for requests to hook into services like Facebook. Getting a complete solution for these was unfortunately beyond the timescale of the Friend View project however.

Note: If you still want to lifecast where you are and what you are up to with your Facebook friends, you can try it using Ovi: Learn more.

“Routinely” – A visualization based on some Friend View data

Play from YouTube

Routinely is an experimental visualization built to abstractly present global activity. The project was motivated by an observation that many users update their friends about their current activity and location. Routinely considered some common activities such as working, eating, recreation, etc.. Aggregating messages by activity, time and location, gives a small glimpse of the range of activities people are engage in, all around the world.

Anonymized user data was used as the input for this visualization. About 3000 status messages are used as the basis for this visualization; these were anonymized status messages that contained a location (explicitly added by the user) and classified as matching against certain English language terms.

• WORK (Shown in Red) - Terms: Work, School, Class, Office, Meeting
• HOME (Shown in Orange) - Terms: Home
• RECREATION (Shown in Blue) - Terms: Party, Shopping, Mall, Store, Gym, Play, Movie
• EAT (Shown in Green) - Terms: Beer, Pizza, Dinner, Coffee, Lunch, Breakfast, Eat, Food, Drink, Pub, Bar, Restaurant, Dining, Hotel, Cafe
• SLEEP (Shown in Light Grey) - Terms: Sleep, Night
• WAKE (Shown in Dark Grey) - Terms: Wake

The looping visualization is compressed into one 24-hour period, using the time-of-day of status messages. Markers on the map indicate prevalence of a certain kind of activity for the location it is placed on. The color and size of a marker indicate the category and relative amount of that activity (multiple matches packed together). At the bottom of the visualization, a stacked chart presents the activities predominant in that timezone. Time-of-day for any location is shown in the timeline above the visualization. Additionally, the day/night overlay placed on the map makes it easy to discern time-of-day for any location.

The visualization helps give an idea of where Friend View was used for some common types of activities. It is intended only to give a small interesting glimpse onto some typical types of usage patterns. The “Routinely” visualization was created by Srikanth Narayan from the Data Insight Program of Nokia Research Center, Palo Alto.

Last word

Friend View was a very interesting project in a number of different ways. Having the opportunity to serve the Beta Labs community, and getting your feedback was one of the most rewarding aspects for the Friend View team, and we thank you! Hopefully this final blog post is interesting feedback for you.

Best regards,

James Reilly
Friend View project manager