What are Last.fm Friends, Part 1

As regular readers know, I have been working on a survey about the nature of “friending” on Last.fm. The study is motivated by a few concerns:

Developers have given us the term “friend” to use, yet it often seems a poor fit. How do people within these relationships define them?

What is the quality of these relationships?

What motivates these friendships and, in the context of a fan-based site like Last.fm, what role does shared taste (in this case musical) play in forming and maintaining these relationships?

    I’ve now completed the survey, and my former Ph.D. student and now faculty member at Ohio University, Andrew Ledbetter, and I have started crunching in order to figure these things out. Here’s some early descriptive findings for the curious:

    After data cleaning (eliminating partial surveys, eliminating minors’ surveys, eliminating those with no friends — I’ll use the no-friends data in qualitative analyses), there were responses from 559 people in 47 countries who had at least one friend on Last.fm. People were asked to describe their relationship with the first friend that appeared when they opened their profile page (which is random).

    THERE ARE MANY CROSS-SEX FRIENDSHIPS ON LAST.FM:

    Most (51.6%) of the friendships are between people of the same-sex, though there were many (46.6%) cross-sex friendships. Though it’s hard to get real numbers on this, it’s certainly higher than the ‘offline’ norm.

    THOUGH THERE ARE FRIENDSHIPS THAT SPAN BIG AGE DIFFERENCES, MOST ARE THE SAME AGE:

    The mean age difference between friends was 0.89 years.

    MORE LAST.FM FRIENDSHIPS ARE INTERNATIONAL THAN LOCAL:

    40.4% of friends live in different countries. 25.4% in the same country. 10% in the same part of the country. 16% in the same town, and 5% in the same neighborhood. In a few cases, the internationalism was part of the friendship’s appeal. For instance, a handful said that the chance to practice a foreign language was important in their decision to friend the person.

    MANY LAST.FM FRIENDS HAVE MET FACE-TO-FACE:

    41% of Last.fm friends have met face-to-face. 9% of those met only once, 14% more than once but not often, 17% used to see one another but don’t anymore, 40.5% see one another regularly, and 7.5% see one another all the time. That said, only 47% of Last.fm friends first met through the site. The rest had pre-existing connections either offline or in other online sites (fan boards, for example). It would be really interesting to know how this compares with other sites. It suggests that though there are new relationships being formed through the site, to a very large extent it is providing an additional means for people in pre-existing relationships to keep in touch. I am reminded of danah boyd’s argument that these should be called “Social Network Sites” rather than “Social NetworkING Sites” because there is so little new relationship formation happening through them. That said, 47% is no small percentage, so there are clearly new connections being formed through the site.

    MANY WHO HAVE NOT MET WOULD LIKE TO, THOUGH SOME HAVE NO INTEREST:

    Of the 59% who have not met face-to-face, 59% would meet their friend were it convenient, 7% would go out of their way to meet their friend, 7% have plans to meet their friend, and 15.6% are not interested in meeting their friend.

    FOR THOSE WHO USE THEM, THERE IS A LOT OF OVERLAP BETWEEN LAST.FM FRIENDS AND FRIENDS ON OTHER SOCIAL NETWORKING SITES:

    66% use other social networking sites such as MySpace or Facebook, 34.2 don’t.

    Of those who use other social networking sites, 51% of the Last.fm friends are also friends on another site: 18.2% of all respondents with friends were friends on one other SNS, 10.2% on two, 2.5% on 3, and a few participants reported being friends with their Last.fm friend on four or more social networking sites. This strongly resonates with the claim I made in my recent paper on Swedish indie fandom online about how people are spreading their relationships across multiple sites.

    I have several thoughts about this — I expected more people to have met through Last.fm, I expected fewer to have met face-to-face. I was surprised and pleased to see the number of international friendships. I was disappointed to see the age-similarity of the friendships (perhaps because my friends are on average considerably younger than I). Even given the age-range of users, less than a year’s average distance suggests people are not generally reaching out across generations via the site.

    I am very interested to hear any thoughts readers might have on what here is surprising, obvious, connects with things you’ve seen going on elsewhere, etc.

    Many more findings to come as the more complicated analyses get going. Stay tuned.

    TwitterFacebookDeliciousFriendFeedLiveJournalStumbleUponDiggLinkedInMySpaceTechnorati FavoritesShare

    Comments (4) to “What are Last.fm Friends, Part 1”

    1. That’s very interesting. I guess the fact that most people have met in real life might also explain the small age difference between friends? Btw, have you followed the discussion on open social networks?
      http://bradfitz.com/social-graph-problem/

      I’m looking forward to Part 2! :-)

    2. Regarding the cross-sex friendships, I’m guessing there are more men using Last FM than women. (Every one of the music forums I’ve been on have had decidedly more male than female members.) I’d say I’ve become friends with more men than women through Last FM. I wonder if more women claim to have cross-sex friendships than men.

    3. Kathy — Great question. There were more male respondents than female, and I’d agree with your assessment that Last.fm seems to have more male than female user. And in answer to your specific question == Yes: both women and men are more likely to be friends with men than with women.

      I thought women would be likely to have more friends total than men, but that is not the case.

    4. With regard to the age similarity, musical tastes change with the generations. It would be interesting to see if those who formed friendships with large age gaps had music genres in common and similarly if those who did not perhaps chose music within a narrow range of what was/is popular in their youth. How much musical choice influence the friendship?