The “Friends of Friends” experiment on Orkut

I admit it, I sent out an email message to about 1500 people on Orkut to see what would happen. To explain why, let me start by saying that the more I think about social networks, the more I find that the concept hasn’t been thought out by anyone yet, whether it’s the Ryze network event in Denver next week or whether it’s the Orkut “friends of friends” communication capability.

But on to my experiment…

On Orkut I have about 60 friends on my list now, including some folk that have really come out of the woodwork, including a woman with whom I went out on one date about, oh, 18 years ago? Each of them has friends, and my overall network of friends of friends is about 1500 people (which means that each of my Orkut friends has an average of 25 friends. Multiplication is a wonderful thing).

Here’s the message I sent out:

Alright, Orkutters (Orkutees? Orkutathons?), how about a few links to
my new 404 error page site? If you could find some excuse to blog me
or otherwise point some folk at:

It’d be an interesting experiment AND I’d appreciate it too.

Thanks, y’all.

Dave Taylor

The responses were quite interesting, starting with the inevitable “you’re spamming me”, though when I engaged that writer and asked what wouldn’t be construed as spam in his book, he wrote back “It’s true that it’s hard to think of things which would interest all the friends of your friends which can number in the hundreds, but I’m more tolerant of the ones which 1) are of somewhat general interest and 2) offer something of possible value, either information or an opportunity, instead of asking for something.” So that’s one data point: offering is better than asking. Maybe.

The next response was a bit more uplifting:

Please say this is the Dave Taylor from a hundred years ago working on shopping malls? I still have your book, if so, with ‘tune in, turn on, burn out’ inside. It has followed me from the InterNIC, to Level 3, to ONYX, to Carrier1 to SBCIS <G>

If not, then lucky you to have the name of an internet rockstar. Cheers, -ren

I’ve never thought of myself as a rockstar, but, well, that is me.

Then another positive response:


This reminds me of a story from my days. My favorite question to ask new people working there was what’s the most important page @ Everyone would say the front page. WRONG. The correct answer was the 404 page… Since the vast majority of their traffic was to their web hosted sites, there would be tons of 404 traffic from them. All their highest priced ads were on the 404 page. When the needed some cash, all they had to do was tell them pr0n checkers to kill some sites.


I’ll help. How about a link back to my page (PR3)?


but in order for google to get it to work properly
you need the text in the anchor to be something no?
BTW.. Iinked it on my home page which google sees a bit

There were more in this vein, but if you’re keeping track, that’s three people who linked out of five who responded.

More challenging to figure out is that at almost exactly the same second, I got the following two messages. First, one that said:

This is terrific – made me laugh out loud at 6.00 am in the morning (sad soul that I am). I’ll do a short news piece on this in my Linking Matters newsletter which will go out next Wed. I’ll also link to you from our site.

That was a real bit of value from Orkut – thanks a lot.

and then another, the most blunt and crude of the lot:

Lovely. Spam here already.

You truly must be the author of
one of the Dummies guides if you
can possibly think it’s ok to spam
random people with your drivel. With
any luck at all, you will never be
presented with the chance to
reproduce, so we can at least
hope that subsequent generations
will be spared from any repeats
of this insanely inane behaviour.

Quite disappointingly, this person has not responded to a polite message I sent him asking for a clarification of what he felt would be appropriate friends of friends communication. I can only hope that he will respond so we can engage in a dialog. But I haven’t had the heart to tell him that, yes, I have reproduced! 🙂

Meanwhile, another message arrived while I was contemplating that last one:

I just noticed my mail messages on Orkut; I don’t get here often but they may have to change. Anyway, I like your effort and it makes me realize that a sociologist type of thinker could look at absolutely anything and find a topic worthy of research. I don’t have plans to blog this particular sociology-of-the-Web topic now, but I’ll keep it in mind; often I find novel ways to work in topics.

and then

I added a link, at But don’t expect a torrent of traffic all of a sudden or anything from my humble blog.

But of all the messages I’ve seen, perhaps the most amusing is that someone else has also sent out a “friends of friends message”:

In the spirit of Dave Taylor’s recent request for links and visits to his site,, I’m taking this chance to point you — all of you — to my web log,

What I find the most interesting about this entire experiment is that in a new social venue there aren’t any established norms for behavior so any behavior at all is quickly judged as acceptable or not acceptable by the community. But here we see a complete lack of consensus. Also worth noting is that not one of these messages came from a “friend” — they all came from “friends of friends” — and that about a half-dozen long-lost friends have shown up and requested a connection subsequent to my original message being sent.

And yes, I know spam. My inbox overflows with spam from fictitious people and bogus companies. Hundreds of messages a day pour in. But I have a pretty typical sense of what is and isn’t unsolicited commercial email and while a chance to increase my anatomy or opportunity to refinance my house clearly crosses the line when it’s in my inbox, does the same offer from someone I know, or a friend of one of my pals similarly violate behavorial mores?

For that matter – and thanks to my pal Jodie – another question arises: if you don’t want to expand your horizons to receive communications of whatever sort from your “friends of friends”, why would you join a network like Orkut in the first place?

I’d be very interested in your reaction, dear reader, to this entire experiment and the resultant communications!

3 comments on “The “Friends of Friends” experiment on Orkut

  1. And another message just arrived:
    “Looks good to me, and I’d be happy to link to it. Please submit this to my webmaster resources directory at”
    I’ve also received private communication from at least one other Orkut member who is similarly baffled by what is or isn’t an appropriate use of the “friends of friends” capability: she sent out an offer of free publicity for some potential partners and received ten queries of interest and two accusations of spam.

  2. Your comments on getting replies not from friends but from “friends of friends” is interesting. Ties in with the findings of Mark Granovetter in his paper, “The Strength of Weak Ties” (weak ties = friends of friends, strong ties = friends) which suggests that when it comes to finding a job, getting news, launching a restaurant, or spreading the latest fad, our weak social ties are more important than our cherished strong friendships. (Quote from ‘Linked’ by Albert-Laszlo Barabasi).
    The thinking runs that your good friends are bad for spreading news because they form a tight cluster where everybody knows everybody else. Your friends of friends however are members of clusters that you have no contact with and therefore they’re great for spreading news.

  3. Very interesting, Ken. I think you’re correct. I’ve been thinking a lot about these networks and I surmise that my ‘friends of friends’ who are, if you will, ‘friends waiting to be met’ are good connections, but there are also friends of friends who are definitely not sympatico. Which is why each of us has our own – hopefully unique – social circle. Modeling that with a simple computer program is perhaps not as easy as these sites suggest…

Leave a Reply

Your email address will not be published. Required fields are marked *