Friday, August 10, 2007

networking on the net

myspace & facebook were, i think the pioneers in social networking websites. Sites which link people to people. And then tons of other websites followed. Some for social networking and some for professional networking. I myself use orkut for social networking and linkedin for professional networking.

First of all, we would go a bit in depth about the technicality of these web sites. Well basically in terms of software engineering - all networking websites are graphs. Each node having number of child nodes and so on. The best way to built such sites is to have a linked list type of structure where each user or node and its corresponding links are stored in memory.

The architecture for such a site could be easily designed for a small number of users. All you got to do is store the information that user "A" is linked directly to users "B", "X" and "Y". And that information could be stored in the database. But, as the site grows, this information multiplies exponentially. It is like 2 users invite 4 others and those 4 would then invite 16 others and so on.

It is when the data reaches a critical mass, that extracting information from the site becomes slow and heavy. This is where the system starts to smoke and the site goes slow. The way to solve this is to have the information - which is in the form of a graph in the memory. Where also it is easier to traverse and get information whenever needed.

Now lets look at it from a business perspective. What do entrepreneurs get out of building and maintaining such heavy sites? Lets look at social networking websites first.

You have the benefit of searching and getting in touch with long lost friends. And you build up your network. The way to get revenue out of this is firstly ads. Now social networking sites are places where you stick and explore. Of course, i would spend more time on orkut than i would spend on yahoo or gmail or any other website. So, if a user spends more time of a website, it is more likely that he would click on one or more ads. And this is where the revenue is. But is this revenue good enough? I cant comment on that. But look at it from this point of view, that the information available from these sites could be used to generate very relevant ads. And the ads could be linked from user to user. So if i click on a particular ad, then people of similar age and sex who are in my network would be interested in that ad. At least it increases the relevance of ads shown.

And what about professional networking sites. In addition to ads, professional networking site is a bit more serious area than a social networking site. So, if you are on a professional networking site and you have got your cv over there, then someone who is looking for a similar kind of profile - might find you in his network and ask for reference. So in addition to getting a person of matching profile, you also feel more secure about the guy - cause he is known by someone whom you know. And that someone can vouch for his work profile. So, the revenue model in professional networking sites is to charge for contacting people. So there is someone who is not in my direct network and i being an HR person find his profile interesting, then i will have to pay to contact him. It could be looked as a next level of job-site, which could be used to fill up "top level" positions. People on the top who generally are not on the normal job-sites.

But what i fail to understand is that is this revenue good enough. Well, only time would tell. Ofcourse, once the site grows huge with trillions of users, the revenue can be substantial.

I like the model of techtribe, which allows a user to post a job and another user to refer someone for the job. It is more focused on the idea of head-hunting using professional networking.

And the latest addition to this list is brijj from info edge ('s parent organization).

Maybe a day would come when you would look into your network to find a good plumber you could trust.

P.S. List_of_networking_websites

No comments: