My Gmail inbox is really a catch-all place for communications. Its composition includes transient and long-term relationships, business and personal communications, and a host of other things that I file away for later review. I tend to treat my gmail Inbox like a big junk searchable junk drawer -it’s a good place to keep stuff for later use.

That being said, if you wanted to bootstrap a social graph from my inbox behavior, you would need to understand a few things:

Context – For any given person with whom i communicate, what’s the context? Are we friends? Colleagues? Business associates? Alums from the same school or program? All of this matters as it determines what I might want to share with them.

Flow – What’s the direction of the communication? How often do we communicate? Is it unidirectional or bidirectional? Take Amazon, for example. I receive and read a lot of email from Amazon. But I nee wrote back. I also have people in my inbox who send me lots of interesting stuff, but it’s mostly a broadcast relationship. Raw communications volume is not necessarily a good proxy.

The two issues above can easily be solved by making a system opt-in. That alleviates the need for Google to try to infer context and flow as I can do that heavy lifting for the system by choosing which people to include and invite.

The reality, though, is that Google is a bit late to this party. Let’s dissect the situation. Google’s best and biggest social graph (or the biggest and best social graph that they can access and index) is Gmail. They can’t get at Facebook or Twitter, so they have to work with what they have. While an opt-in system, would have been more privacy-friendly, it almost certainly would slow adoption and growth. Inviting people to something new is time consuming. And a service like Buzz isn’t interesting or useful until you have enough people on it, sharing information, and engaging to make it feel vibrant. So if Google wanted to try to catch up in social content sharing, the fastest path would be to put it in Gmail and make it default opt-in.

I’m also waiting for Google to start looking at what I’m doing in Gmail and starting to insert contextual Buzz prompts. For example, how long will it be before I’m sending out a link to a cool YouTube video to a group of friends and I get a prompt to post it to buzz in addition? The same is true of Picasa photos, Flickr photos, or any thee piece of web content. Gmail Labs already has some handlers that can detect and display certain types of content – why not apply those to Buzz?

In th end, Buzz will work of it drives traffic to content sites or helps people engage around content.

What am I missing? Can someone help me out?

My Setup
I have a fairly atypical setup. I have several thousand contacts, 8ish calendars, and a lot of other information that I like to have available across all of my computers and mobile devices. I have 4 Macs (1 Mac Mini, 1 Macbook Air, 1 Macbook, 1 Macbook Pro – yes, that’s ridiculous and I accept that), 1 iPhone 3G, and 1 iPod Touch. I have two core requirements:

1. Support for OTA sync for my 8 Google Calendars and 1 Gmail account
2. Ability to differentiate between what information I’d like on my PC and what information I’d like on my iPhone

Overall, I found the revised MobileMe to be very unsatisfying for me edge case needs. A few things that didn’t work well for me:

No support for contact group syncing – I try to keep up with a lot of people. One of the key ways I make sense of all of my contacts is by putting them into groups. I don’t necessarily want to sync all groups to all devices at all times. For example, there’s a relatively small set of contacts that I like to keep on my phone and a fairly large group I like to keep in my address book. As far as I can tell, MobileMe does not support this use case. It’s an important one for me, even if I am an edge case.

No support for my primary email address(es) – While I’m sure I could figure out a way to make forwarding work, I don’t want to jump through hoops to support my primary email addresses (my personal domain and my gmail account). An OTA push email system that doesn’t support my primary email addresses is a non-starter. I can’t believe I’m alone here.

No need for the Gallery or iDisk products – I don’t have much of a need for the Gallery product and I already have a preferred web-based storage product (Dropbox). Neither of those features are of particular value to me and I don’t see a need to pay for them.

No substantive value above and beyond existing OTA calendar sync products – I’ve used Nuevasync and Google Sync to keep my Google Calendar in sync with my various computers and devices. Both products work well – I prefer Nuevasync over GCS because it allows me to sync more calendars and that’s the one I used first. MobileMe’s calendar product works well, but not meaningfully better than the Google or Nuevasync product. I don’t see any reason to pay when the free alternative is solid.

I have been pretty happy with NuevaSync but am tempted to try out the Google product when and if they release OTA mail sync and support for more than 5 calendars.

Increasingly, I find a lot of my friends who are using Facebook as the place where they keep their contact info up to date. One solution would be for Facebook to open up and let other services tap into that data store to keep their address books up to date. That’s geek-cool, but I’m not sure how many people would want to use it.

What’s cooler, and potentially more powerful, would be to have a Facebook address book application that works on my phone. What I really want to be able to do is “click to call” any of my Facebook friends right from my Blackberry. Instead of having to keep track of their contact information in a separate app, I’d like an app that ties my Facebook friends and their contact info together in a seamless way.

This shouldn’t be that hard to implement. A user could have a simple configuration option where all click-to-call requests from “friends” would be routed to a number of their choosing (cell, mobile, voip, etc).

I think this would be a great way for Facebook to continue to extend their presence on mobile phones and reinforce the value of keeping your contact info up-to-date on Facebook.

2. None of the social networks have really launched a great desktop IM application. I remember when MySpace launched IM and the product was not that compelling. I can’t remember any other social network of scale launching its own standalone IM client. Arguments about downloadable clients aside, I wonder if there aren’t other reasons none of the other major social networks has launched an IM client of its own.

3. I’m surprised that one of the multi-headed IM client guys haven’t done the same thing. When I first saw social.im, my immediate thought was that was something I would have expected the Meebo or Trillian guys to do – create some nifty FB app that would allow you to pull your Facebook friends into your existing cross-network IM experience. Perhaps they have something similar in the works – it would be a good way to at least continue to seed / grow a user’s IM network. Those guys are already in the business of bridging networks so sucking down your FB friends seems like a logical next step.

On that same note, I’m not surprised that none of the major networks has launched such a product. The real value in something such as social.im or imo.im is that it’s IM network agnostic – I don’t need to have my friends self-identify as users of a given service. All I need to know is that they’re on Facebook.

UPDATED – I am all out of invites. If I get more, I will work my way down the list of people in order of comment timestamp.

1. For most people, this is a non-issue. I know a lot of people who actively publish the URL and location of their shared items feed. I even keep mine on my blog, including the most recent items I’ve starred. For most people, having others see the items they’ve shared is not a big deal. If you really want to keep track of stuff and not share it with the world, just create your own tag – I use the tag “followup” for interesting stuff I find but don’t want to share with the world.

Also, it’s not as if the GR team really “exposed” any private data. Your shared items feed was always publicly available – it was merely obfuscated with a long and gnarly URL.

2. “Passively social” products are going to encounter this problem time and time again. The advantage of passively social products is that they don’t actually require you to “friend up” your network to get started. In an ideal world, a passively social product will be able to infer or distill the appropriate social network with which to share your information. We’re clearly not there yet, so the only thing we’re going to get from these first generation products is a very coarse set of relationship data. Knowing that I’ve communicated with someone on Gmail is a very coarse signal – it’s better than nothing, but not enough information on which to make a high-quality decision.

As excited as I am about passively social products, it’s going to take awhile to be able to match the user experience associated with actively social products like Facebook and MySpace. Without explicitly declared (and verified) relationships, passively social products are going to need multiple signals to make good decisions about the circle of people with whom I’d like to share data. For example, if a passively social product had access to my IM, phone, email, and calendar logs, I bet it could make some really good decisions about the folks with whom I communicate and the strength of those connections. There isn’t anyone who has that information today across all those platforms. As a user, I don’t have it myself and couldn’t even turn it over to a trusted 3rd party to analyze it even if I wanted to do so. And I can’t think of a single company I trust enough to provide all of that information. I would maybe trust Google, but I’m not sure – that’s a lot of information and power to turn over to a single organization.

Tags: googlereader, web20, privacy, google, opensocial, social networking

Overall, I do agree that email inboxes do contain a lot of interesting data about people and how frequently they communicate over email and potentially IM if a vendor offers both products in an integrated fashion. That being said, I don’t see how any of the top web email providers (Microsoft, Yahoo, and Google) can use this email information to build new social networking products. There is, however, an opportunity to use that data to power other people’s applications.

What additional value would I get in using one of these systems over Facebook, MySpace, or my current social network of choice? Details on these products are sketchy at best. However, almost any social networking product worth its salt has a contact importer. Once a user imports his/her contacts, he or she can then determine who from that subset of people he/she would like to invite. Is having a machine prompt to do this for personal social networking of great value? I can see the utility of this auto-population or auto-discovery in a work context (Xobni does do a good job of showing me my own correspondence patterns and I can imagine many things you could build on top of that data – the work use case is different as I think work communication patterns tend to be more dynamic than personal ones). Nothing I’ve heard in the limited details that have come out gives me reason to think that they’re on to something bigger.

I’d also say that if “powering up” this network requires me to create a new profile page, it’s a non-started. I’m out of that business for now unless or until I see a really great application that’s worth the time.

Webmail inboxes are a mess – I have yet to use an email product that has an even decent address book. All of the email address book offerings from the Big 3 email providers feel really dated. For example, the Gmail address book does not do a very good job of de-duplicating contacts. I have folks in my address book who have multiple entries and I’m not interested in going through to manually de-duplicate them; I’m counting on a machine to do that for me.

The larger point here is that I don’t know how you can build a really good, effective social networking product on top of email if you don’t do something to put some good, quality structure around the data. Social networking services who are sucking up email addresses to match a user’s inbox with their database of contacts don’t have the same problem – you just throw away the ones that don’t match (or allow a user to invite them). It’s a very different situation if you want to build a whole new social network product with email as the foundation.

Cross-functional collaboration is not easy inside of large companies – This is a fairly obvious point, but big companies are notorious for having internal challenges when it comes to cross-product collaboration. When one of the products in question is email, I don’t imagine that will be an easy conversation – nobody wants to play around with an interface that touches tens or hundreds of millions worldwide.

Think about the refresh cycles for webmail systems. How often do Gmail, Yahoo Mail, or Hotmal get updated? Not that often, and I have to think that touching those interfaces requires a lot of signoff and a strong conviction that the proposed changes will positively impact a wide number of people. Otherwise, you might end up with angry users. I have a hard time seeing any of these companies acting aggressively with one of their web crown jewels.

The end game ought to be to make this information available to other services and make mail the data platform, not build new applications. Sadly, I don’t think that’s a very interesting business to be in – I don’t know how you re-establish yourself as a major player in social networking by simply providing the data layer that powers other applications.

So what’s next? Well, I think what’s next (and by far most interesting) is some concept of the “strength” of a connection. Specifically, today I can see a lot of my friends’ social networks, but I have no idea for the relative strength of connections. Sure, if I see Person A knows Person B, I can always make an offline inquiry to see if that connection is strong or weak. But very soon I think we are going to have tools like Xobni that profile communications patterns and surface that information both to end users and to other applications. And it won’t be just social networking and community applications that benefit. Enterprise applications (collaboration tools, CRM tools, HR/recruiting systems, etc) will all benefit from having access to some of this information. We’ll call this contextual “strength” Phase III.

Phase III is really interesting to me because I think it has to be a largely machine-driven approach. Communication patterns are too dynamic for any user to bother continually updating “strength” of connections. Also, as Xobni has shown me, if you are a power emailer you’re likely to be surprised by who shows up as ranking highly. There’s no reason the same can’t be done for IM. I’m not sure that I’m going to turn my phone logs over to some 3rd party analytics company, but IM and email would be a pretty decent picture of what I do and with whom I communicate. Passive profiling of communications patterns is going to be really interesting and I think will expose really interesting information about the nature of communications. I think Xobni is on to something really cool and big as it’s delivering value to me today (even though I have to use it in Outlook) and I can see a path to a lot more value in the future.
As an aside, I think this is the best shot that Google, Microsoft, and Yahoo have to wedge their way back into social networking relevance. They already own the message stream and have the data they need to get a sense for who knows whom. It will be interesting to see whether they choose to open this information up and let other applications take advantage of it or whether they use it for the bedrock of their own auto-generated social networks.

• The interface is quick – The Gmail interface is really snappy. It loads quickly and refreshes automatically. As is the case with most Google products, speed is clearly a high priority and it works really well.
• I can process/triage messages very quickly – Once you master the key keyboard shortcuts (j,k,x,n, and p are the crucial keyboard shortcuts to master in my opinion), it’s really easy to blast through a bunch of messages in your inbox and quickly triage them or otherwise mark them for future reading/evaluation.
• Search trumps foldering once you make the big leap – I am an active labeler in Gmail, but I’ve gotten lazy/sloppy with some of my labels over timer. The key “ah ha” moment in most Gmail user’s evolution is the moment in which the light bulb goes off and you realize that search is a more powerful navigation paradigm for email than foldering. It’s a leap of faith until you make the change.

One thing worth noting about the three things that I like most about Gmail – they only really matter if you are in a position where you need to manage (read, write, retrieve, and share) large amounts of email. If you don’t get a lot of email, most of these features and benefits don’t buy you much. In fact, you probably would not even bother mastering these things if you are not a power emailer.

The subject of to whom Gmail-like interfaces appeal is a subject for its own post. For the purposes of this post, we’ll just say that things like Gmail are designed for expert power users and have a high bar to adoption. However, once you adopt, it’s hard to switch.
After trying to go back to Outlook as my every day mail client, I’m finding it to be a very difficult adjustment. There are 3 things that stand out after a few days of non-Gmail email existence:

1. I miss keyboard shortcuts – a lot. After being out of the Outlook experience for over two years, I find the keyboard shortcuts to be a bit slow. Also, with the myriad number of formatting and presentation options that Outlook offers, it’s hard for me to actually remember all of the keyboard shortcuts at my disposal. The relatively short list of Gmail shortcuts that I can use really do the trick and allow me to get my work done quickly.

2. The Outlook client feels really slow. The time spent selecting messages, opening them, waiting for them to open, and then closing them and moving on to the next message just feels a lot slower in Outlook. I don’t know whether it’s because I’m less familiar with the interface or if there is something else at work. It just feels like it requires more keystrokes and work to get through my email using Outlook.

3. Working offline is not as big of a bang as I had thought originally. Now that I have a broadband access card for my laptop, there is relatively little time when I cannot be online if need be. Suddenly the ability to use my mail client offline is of less value as I find myself offline with declining frequency.

At the end of the day, I think that has big implications for folks who are building apps that plug into Outlook. Sure, not everyone in the world who uses email will switch from Outlook to Gmail. However, I know a lot of power email users who are moving away from Outlook and adopting Gmail. Given that so many of these interesting email products target power users, I am curious to hear how they will deliver their solutions to the (relatively) closed environments that we find in webmail. Greasemonkey scripts? Biz dev deals to get access to the platform? Lobbying hard for more opennes and APIs? I’m sure they’re all thinking about this stuff and what it means for their businesses.

As always, comments are welcome.