The New York Times is summarizing some interesting plans from Google and Yahoo to turn their email systems into forms of social networking. This idea could have a lot of potential, as the Yahoo’s Brad Garlinghouse points out in the article that Yahoo has a lot of information about an individual’s social relationships – for example who they email regularly – and this info is simply begging to be mined to help users navigate their increasingly complex online worlds.
Joe Duck
Have Blog. Will Travel. Vaccinate!
Joe Duck 
Email is the next natural integration into social networks. The next round of social network adopters have all adopted email and are ripe for social networks.
Contact lists can be merged with social network relationships to save the inbox.
definitely its a great idea — and it will work
Definitely agree with mitesh. In my business/social network, I am shocked at how many of my colleagues and friends are participating in LinkedIn, and although it obviously does not use the same mechanism we’re talking about here, the fact that I can very easily find folks I know on LinkedIn and compile a quick bunch of personalized emails (all in one interface) makes me much more inclined to use it. If it were truly linked to my personal emails in an intelligent way, all the more so.
If (or When) this happens, its going to make a huge splash.
Gee, I thought people were not even emailing anymore…just twittering each other or posting on blogs and replying to other blog posts.
Sure Google “reads” subject lines in emails and pushes ads that supposedly comport with the subject lines, and other major portals want to focus on contact lists. My contact list is almost ten years out of date. And many of the people who email me frequently I don’t particularly want to hear from. Indeed many of the emails I receive, I delete without opening, much less reading, them.
Is this some technique to discern from emails what sort of a “lapel pin” the sender would wear (to continue my analogy from an early posting)? Well, the small chessboard lapel pin that might be worn by some people on some occasions to indicate their hobby to others might be a large and prominent one in some situations. Email might reveal this if a certain topic was frequent. I think data quality issues would mandate ‘self declared’ attributes rather than ‘algorithmically derived attributes. (Checking a box that says ‘chessplayer’ rather than a subject line in an email frequently reading ‘chess game’).
What I’d like to see is a *good* algorithm that makes logical grouping suggestions on your lapel pin concept; one that is 80-90% accurate and makes it clear that this is a suggestion, but needs your attention/approval if you want it to be truly correct. It seems like most apps today are either all or nothing in this regard – they either abandon the idea of helping you with this altogether or they force you to use what their algorithm suggests…I know its a difficult problem to solve, but seems like someone in Google’s white tower could get it right. Or at least close.
Anyone else having a different/better experience?
I used the ‘lapel pin’ analogy because I thought that social networking would best be accomplished via a variety of such electronic lapel pins of varying strength. The problem that I posed before was how to make such lapel pins be selective in their strength as well as precise in their meaning. We now have zillions of sites that lay claim to being ‘social networking’ sites. Success will go to whoever creates the best matchmaking algorithms. Sort of the way a successful hostess is one who consistently seats interesting guests next to each other at a dinner party.
80-90% accurate
Yes, you’d think this would not be too high of a standard.
FG agree that “matchmaking” is a key here. My take is that Open Social will lead to all of us exposing a lot of our details which in turn will lead to a *portable* social identity. I don’t think any site can or should be the key place to meet – rather we should all have our own identity package that travels with us as we surf and to which we control access as much as possible.
The portable identity package is the right idea for sure, but it will have to have different “views” at different sites. Different granularity for different situations. After all, at a cocktail party a famous author might just want to be known as ‘author’ or as ‘neighbor’ or as ‘gardener’ or ‘single’ or ‘wine-lover’ or ‘auto mechanic’.
The portable social identity concept is really interesting…hadn’t heard it in terms of the social network concept (I mean, its there already in a sense obviously, but very cool point Joe), but that sure would change the social landscape if you virtually walked into a site and folks “recognized” you from other sites based on things other than a common handle. . .not sure where that is going, but you have my brain doing little joyous dolphin-like flip-flops trying to grok what that means.
A good number of people are already recognized (with some degree of certainty) based upon posting styles, vocabulary, stylistic habits, etc. Sometimes a favorite topic arises and a posters comments reveal his identity, such as conversation turning to Saab automobiles or something.
Thx Metro!
FG noted: (Portable identity) will have to have different “views” at different sites.
I see this as the challenge, especially because once the cat is out of the bag in terms of your identity you can’t easily put it back in. So if somebody goes to an “open” health forum and mentions that they have several in their family with cancer and then they allow people to see their picture and identity, how can we make sure this is not used to deny them health care or used to find out about their family. Generally I think the answer is “we cannot do this” and what we need are rules for the *use* of confidential info rather than all the rules now floating around about how it gets distributed.
I think Fools Gold got it… Whoever is able to harness the algorithm and capitalize on it will be successful!
10.A good number of people are already recognized (with some degree of certainty) based upon posting styles, vocabulary, stylistic habits, etc. Sometimes a favorite topic arises and a posters comments reveal his identity, such as conversation turning to Saab automobiles or something.
A good number of people are already recognized (with some degree of certainty) based upon posting styles, vocabulary, stylistic habits, etc. Sometimes a favorite topic arises and a posters comments reveal his identity, such as conversation turning to Saab automobiles or something