This last post is an extra bit I thought I'd share. I thought it would be interesting to do a more forward looking post that postulates some potential directions that Social Networking could take in the future. This is mostly a combination of my own thinking as well as conversations with industry experts, such as my brother, who is a Senior Director in Corporate Strategy and Acquisitions at Microsoft.
After discussing with him some of the key components for what is to become of current social networks I have found that many of them use crowdsourced information from social networking sites and other online data. I will talk about the following concepts and explain how they work with examples of what they are.
1. Mobile (mobile friend finder using gps, etc.)
Location-based services have been around, and mobile applications such as Google Latitude, Loopt, Foursquare.com are utilizing GPS to identify people's friends and alert their friends of each other's whereabouts. Basically, the GPS data is sent to these programs through service providers and that information is filtered by the programs and sent back to the service to the individual with the smart phone. Also, they will make it much easier to upload photos, videos, etc. to Facebook from your phone. So now people are not only virtually communicating in the virtual and digital world that Facebook is, but people are using data to meet up and actually be physically present with their friends. As we have discussed, there are major privacy concerns and other issues around location-based services, however, in time I believe many of these issues can be resolved and/or made not so harmful."
2. More granular permissions (e.g. easier to share some photos with subgroups, not with others)
This is basically the ability to create extremely detailed ‘contact lists’. Facebook has made this possible so now one can have a ‘college friends’, ‘family’, ‘work friends’, ‘professional contacts’ etc. lists so you can filter your news feed by these lists and change your privacy settings for each. For example, you don’t want coworkers to see your photo albums so you can prevent them from seeing those. Well, in the future these lists are going to be more and more important. The idea is that, in the end game, you can actually have many profiles in one application (Facebook), so you don’t need to have a separate LinkedIn account for professional contacts, etc. You just have different versions of yourself that you expose to your different contacts based on what list they are in. And, by the way, there is another huge advantage to doing this, which is that it makes the News Feed algorithm much smarter. The more data you give the service about the nature of your relationship with your contact, the more easily the algorithm can filter the most relevant stuff. For example, I can even tell my news feed that I want more feeds about ‘friends’, ‘family’ and less about ‘coworkers’, so the feeds I get are more relevant to what I’m looking for. In the past couple hours, CEO Mark Zuckerberg sent out a statement about altering the way Facebook users control their privacy settings. Now that networks have become so large, limiting people from seeing content on your personal profile by choosing social networks is too broad. Facebook is allowing users to modify their privacy setting in this granular sense that I have been talking about. Instead of having you networks have different permissions, you, the user gets to decide what content you would like to share with whom.
3. Extension of your social network into activities beyond the facebook walled garden (e.g. will see what movies your friends are renting on netflix)
This is a reference to Facebook Connect. Facebook is fast becoming the ‘identity platform’ for the Internet, in that they now let you use your Facebook ID to log in to Amazon and 3rd party sites (anyone who is a Facebook Connect member). The reason Amazon uses Facebook Connect is for the following reasons:
A) They can then push feeds about you back into your contacts’ Facebook news feeds (e.g. “Connor just bought this book about honey bees on Amazon.com” shows up in your Facebook contacts’ news feeds.
B) Amazon can suck out your ‘social graph’ information so they know who your contacts on Amazon are, which lets them make better product recommendations to you on Amazon. So basically there will be a ‘social filter’ applied to all your activities outside of the social networks themselves, so that everything you do is a bit smarter. All shopping, entertainment, travel planning, etc. can use your social filter to better personalize their experiences and even tell you what your contacts’ are doing on those sites.
4. Super smart 'news feed' that uses sophisticated algorithms so that we only see the most interesting feeds (e.g. my dad sees feeds related to his stocks, mom to shopping, I see feeds to stanford sports, etc.)
It’s like applying a Google-like super intelligent algorithm to rank the ‘relevance’ of various potential feeds that can be exposed in your news feed. Think about it – if you have 1000 contacts each generating 10 ‘feeds’ (i.e., activities online that are interesting, e.g. (‘Fritz just posted this pic or bought this book on Amazon’) then that’s 10000 feeds / day that the News Feed could show you – which is a massive amount of ‘feed noise’ to filter through. As both the number of contacts grow and the number of ‘feeds’/person / day grows, this problem gets exponentially harder. This is why you need the granular ‘lists’ that provide a lot of ‘context’ on the nature of your relationships with your various contacts. The ‘Feed algorithm’ is going to have to make super hard decisions about what 10 feeds to show out of the tens of thousands that are happening in your ‘social graph’. Not an easy computer science problem. Sounds a lot like Search. Paul Buchheit, the creator of Gmail, started a company called FriendFeed to focus on pulling your social feeds (e.g. Facebook, Twitter, Flickr, YouTube) together and ranking them with an algorithm. He sold it to Facebook. So clearly Facebook gets that this is a huge problem / huge potential benefit to end users and that’s why they bought FriendFeed AND why they have redesigned the homepage again to be more ‘relevance based’ than simply ‘chronological’ (requiring no filters).
5. Better search capabilities / emergence of 'human question and answer' service within a social network (e.g. yahoo answers inside facebook)
The only market that Google has not dominated has been Korea – there is a service there called Naver that dominates ‘search’ instead. Rather than using a super powerful algorithm to filter the billions of websites and rank them, Naver provides Koreans with a great platform for posting questions and having the community answer them. As many people answer questions, over time Naver produces a huge repository of human-generated responses to human-asked questions. Since humans are way smarter than machines (machines are just faster) – the relevance of those answers is way better than what any computer algorithm can produce. So, Google can’t compete because Naver owns this repository of human answers to any question you can think of. It’s likely that Facebook could try and build a similar service. They could make it easy to ask your contacts a question, and then over time if you ask a question that has been answered before you no longer have to wait for the answer – they will just show you the historical answers that were rated the best by the community. Facebook could compete with Google here, which is why Google is scared of Facebook and is creating this program called OpenSocial. Which is basically an attempt to reduce Facebook’s exclusive ownership of / commoditize the ‘social graph’. Google’s greatest fear is Facebook building this repository of ‘answers’ that Google cannot index.
Very thorough and lots of interesting material. It will be interesting to see how these play out and which will be most important (or maybe something new that we aren't even thinking of now).
ReplyDeleteFacebook is a very interesting company because it's so bold (in my opinion, bolder than any other tech company in the last 10 years). Google iterates really slowly. The Google search display page hasn't changed a lot over the years. I think it should, it could be so much richer. But Facebook succesfully went from being simply a directory of your friends to a social network to a real-time feed-based social network like Twitter (but so much more at the same time). The pace of iteration and innovation is just incredible at Facebook.
ReplyDeleteI agree that Facebook is really close to adding location to the news feed. I'm skeptical about more granular permissions at least as far as lists such as "work" and "school" go. There's a lot of overlap, people tagging is tough and without any major user interface innovation, I don't think humans have the capacity to handle the "dispersion" of information (it's hard enough in real life to remember who you told what). I think the trend will be in opposite direction, towards fewer group, perhaps just either "everyone" or a few select individuals (as email, SMS, chat works). The Friendfeed acquistion is a good indication that Facebook is going to integrate with and aggregate other services (it already supports a fair number of services like Blogger, Wordpress, Flickr, etc.).
I think it's interesting that you brought up questions and answers because there's a startup called Quora (still in beta) founded by Facebook's ex-CTO and the Facebook Connect lead that's building a product just as you describe. As Google shows, acquisitions are a lot more likely if the company being acquired was founded by ex-employees of the acquirers.
Perhaps the biggest thing that I feel you missed out is that Facebook is probably going to merge profiles and pages at some point. All relationships in Facebook are symmetrical i.e. you have to confirm someone as a friend but this often doesn't work very well (many people's news feeds are inundated by updates from acquaintances). This is the key difference between Twitter and Facebook right now. Assymetrical relationships will allow Facebook to become more like a real-life social network too.