Ma.gnolia.com, a popular social bookmarking site, added OpenID as an alternative method for new user registration in December, 2006. OpenID is a decentralized user-centric identity architecture that allows a user to use a URL based ID (think of it like a username) to log into a variety of websites.

Earlier this month, Larry Halff, founder of Gnolia Systems shared usage statistics of 5467 users that registered for new service between December 15 and January 31. No specific user data was provided.

Sample Sizing:

• 4800 users enrolled with standard usernames and passwords
• 667 users (12.2%) enrolled using an OpenID Identity URL

The graph to the left shows the percentage of low activity users. 48.9% of all new standard username/password enrollments created accounts but did not "use" the site by adding a bookmark. In comparison, only 45.5% of OpenID users failed to complete a single enrollment.

We have defined "Dormant" users as those who set up an account and bookmarked somewhere between 0 and 9 sites, effectively making them non-active users. In this case, 83.6% of standard registrations fell into this category while only 72.5% of OpenID logins were inactive.

This activity trend shows that though slightly less than 50% of all new users fail to get started using the site, OpenID users were significantly more likely (over 11.1%) to become active users. 

The primary purpose of ma.gnolia.com is to have users  save bookmarks which can then be seen by other users. Users join groups to share bookmarks as well.

Across all users, OpenID users have more than double the amount of bookmarks (80 vs 38). Among non-dormant users, OpenID accounts have on average 289 bookmarks while  standard accounts only have 226.

The third key metric that we were able to track is the overall usage of the site. In this case, we were able to track the number of times that a user added bookmarks. The more days that a user leaves bookmarks, the more active he or she is on the system. As it is shown below, OpenID users were more likely to regularly use the ma.gnolia service. 

As ma.gnolia.com doesn't track logins, it is not possible at this time to determine how often users actually log in to use their account, but tracking bookmarks does provide valuable insight into how different user groups utilize the system.

An analysis of the data shows that a significant number of users, both OpenID and standard, after the initial login load up an initial set of bookmarks that first day, but then become dormant. The dropoff for users that log in a second day and add more bookmarks is steeper for the standard accounts. However, after that, the general usage patterns seem to be similar regardless of how the user initially registered. 

It is possible that there is a correlation between OpenID users and "Web 2.0" services - that people who are adopting OpenID as their preferred method when signing up for new services are more attuned to use social networking sites such as ma.gnolia. As this is an emerging technology, we are still looking to define more data points with other service providers. However, this first data on the uptake and usage of OpenID does show that, at a minimum, the impact of implementing OpenID was positive for ma.gnolia.

This JIT Report is made available under a Creative Commons Attribution-Noncommercial-Share Alike 2.5 License. Please attribute the following: Avery Glasser, activeanalysis.net, Feb. 2007.