Explicit Groups vs Implicit Groups
Kevin Cheng (aka @k), product manager at Twitter and an all around smart guy wrote a great blog post called Can We Ever Digitally Organize Our Friends?. I've been thinking many of the same things that Kevin wrote about since I started to use Google+ a few weeks ago and Kevin's post is a good opportunity to riff on the same ideas. But first, a bit of humor courtesy of someecards:
With that out of the way, here's my thinking on grouping things. I don't like to be that organized personally. I don't file stuff away very well. I never liked folders in outlook. I use a few labels in gmail (about 20) but I label less than one out of every 100 emails I get. I've created two twitter lists (one automatically). I follow two twitter lists. I belong to one or two google groups. I belong to less than ten Facebook groups. I am sure that there are many people out there who are different, who love to organize, file, group, and structure their lives and work. But I'm not one of them.
So when faced with the chore of taking all my friends and colleagues and dropping them in buckets (or circles as it were), I tired of that chore quickly. I stopped at about 20 people. And I am not eager to go back. It is not fun. It is work.
I did create two groups that I think are particularly valuable; my family and our firm. These groups are well defined, they are finite, and they don't change very much. I can see sharing things with these groups in Google+ and I'm happy to have that resource at my finger tips.
But past that it becomes muddy. How about a portfolio company group? Good idea. Except that our portfolio changes a fair bit. We add six to ten companies a year. And these companies add and occasionally drop team members frequently. Am I going to maintain a list of all the people in all of our portfolio companies? No, I am not going to do that.
How about the people I share music and music interests with? That's a passion of mine. But I'm finding new people all the time. I met Tyrone Rubin in a room on Turntable a few weeks ago. I started following him on Tumblr right away. I could have added him to a music circle on Google+ too. But I didn't. If he had an "add me to your music group on Google+" on his Tumblr, I might have clicked on that. But short of that, it becomes too much for me to do such a thing.
The point I am trying to make, which Kevin makes so well in his post, is that friends and interests are not so finite and fixed. They come and go. They are highly fluid and dynamic. And as such, there are very few groups that I want to build explcitly. Family yes. USV colleagues yet. Anything else, no thanks.
This is an oppportunity to use machines. And Google is doing this with Google+. The recommendations on who to add to what circles are amazing. So why make me do the drag and drop thing other than it is fun and cool to do that on a computer? If Google+ knows who my music friends are then just suggest "music friends" when I hit the share button and send it on. Do I care if it goes to a few people who aren't actually my music friends? No I do not. Do I care if a few of my music friends don't get it? Yes, but then I can add them explicitly. I trust Google to do a fine job of this for me. They've proven themselves worthy of the job so many times in my relationship with them over the years. I trust that they can build algorithms like this as well or better than any other company out there.
There's an iPhone app called Katango that apparently does this for Facebook friends. I don't have an iPhone so I have not tried it out. But I hear it is pretty amazing. That's an example of computers doing the work for you so you don't have to do it. That's what I think we should be doing in terms of creating more granularity in our social graphs. I don't want to put my friends into circles. I want a machine to do it for me.
+1 ;)I have a few fairly well defined groups, otherwise everyone goes in a circle entitled “everyone”. I am sure this will kill a lot of G+ functionality for me if/when I start using it consistently, but I simply can’t take the time (or rearrange my brain) to do it “properly”.
Naturally in the real world, we navigate socially guided by implicit signals from our connections, and fine tune our paths as we go.That’s what ultimately I want online with groups and connections and suggestions.
Spot on, imo.To add a bit… Social interactions are influenced by our behavior – which has little to do with meta-grouping. Capturing and understanding behavior, I believe, is the most important signal social networks should be focusing on.Color (yes *that* app) is absolutely on to something in their concept: an elastic network. Their 1st execution was just piss poor.Ephemeral networks have the potential to adapt to our behavior, interests, thoughts – anything imaginable.
AgreePeople expect serendipity on and offline. Finding it (or in Meeps case) feeding it, is where implicit discovery and interest footprints intersect.I blog and think and work on this a bit. I’m #763 on your Beta list. If you want to talk prior, contact me.
I don’t like to file things either, but I’m glad to file myself into things. For example, I subscribe to your RSS feed — I’m filing myself as a “reader of Fred blog.” This is part of why I think Circles are built backward (see https://plus.google.com/103… — it doesn’t really make any sense to me.
Yessssss. That’s exactly right
Thanks. I actually want to build it, but I don’t think there’s a good way to launch a new social network. (Google has the audience built in.) But obviously I’m spending a lot of time thinking about it now — how do I get all my content syndicated to self-identified groups who want that type of content?
I’ve seen this idea from a few and posted about it a few times because I like it so much. I’m guessing / hoping G+ will do it 🙂
Thanks. Unfortunately, one of the comments on my thread is Paul Wilcox, who has a big product role on G+. But I hope you’re right 🙂
Kevin’s three easy steps to launching a new social network:1. make it dead simple to use.2. build it for, and promote it to, a specific niche audience first.3. have the patience to let the network grow, and the passion to make it so.:-D
Are you launching a secret social network?
Shhhhh…. its a secret. ;P
not yet 🙂
“Dead simple” doesn’t mesh well with the huge chicken and egg problem my idea comes with 🙁
get the right ‘product’ person involved and it will…
I have often wondered by there isn’t a modern ‘delicious’ social network…let people write and tag posts (subjot.com anyone?) and instead of following people, you just follow tags or topics and the ‘follow’ idea is flipped to ‘mute’ so that as you are seeing a massive flow of data in the topics you care about, you can pick the people to mute out of your stream that you feel aren’t adding any value…this is the social network that I would want to build (but http://subjot.com is already 1/2 way there IMHO so I’ll leave it to them to build)To me the concept of ‘follow’ is the true thing that’s backwards…I know what I’m interested in, but generally I don’t know who is going to write about it (yet)…and IMHO it’s always easier to identify ‘bad’ than it is to find ‘good’…
i think google has the opportunity to make something like that work very well with sparks, which imo does not work too well right now
I am really liking Google+ it is allowing me to filter out noise in a way that Twitter and Facebook do not. The rate at which I’m adding people onto Google+ far exceeds Twitter because it’s easy to filter people and get a quick view of the different people I interact with. This might be the kick in the ass that makes Twitter do something about it.Circles is going to be a small part of what Google+ is all about.The glaring omissions for me at the moment is an official +1 extension for Chrome (or if there is I haven’t found it), the ability to publically publish groups and the ability to choose what is my main stream.I was really surprised at how good Google+ was on release, I think they have a good product.
Totally agree here. I don’t think Circles are going to become the thrust of G+ (but speaking to a lot of more casual Internet users than I, I may be very, very wrong here). For me, the killer feature is the in-Gmail (and Reader and Search etc.) notifications button. It makes the engagement I get on the product much higher than say on Twitter or on my own blog (if I still had one).
oh lord – that is going to take me a while to figure out. still trying to figure out gmail 🙂
It’s also allowing me to create noise. I stopped posting on FB once I got too many friends because I realized the noise I was creating was irrelevant/annoying to most of my “Friends”. G+ has mostly solved that. I didn’t want to manage too many circles so I made a few important ones. Almost everyone else I know goes into acquaintances. If I don’t know you, you don’t go in. If I want to know you, I follow you. I like to manually put people in circles. It takes less than a second and I sometimes like to put one person in a couple of circles. I would never want Google to accidentally put someone in my “Best Friends” circle. We are finally sharing and socializing in ways that are private and fun. I don’t want a machine to ruin our party. I also really want the option to choose what my main stream is. I would actually like the option of choosing multiple circles as my main stream.
so right about giving you the ability to make noise Kelly and the machine element. Having said that the Google+ algorithm gave me some good suggestions about who I might know/want to follow. Way more accurate than the Twitter one but then that’s probably because I fed it more info.
goole + allows you to filter information according to the list that you create
“But past that it becomes muddy. How about a portfolio company group? Good idea. Except that our portfolio changes a fair bit. We add six to ten companies a year. And these companies add and occasionally drop team members frequently. Am I going to maintain a list of all the people in all of our portfolio companies? No, I am not going to do that.”Solution: Feature for Google+Ability to create invite ‘passcode’ that you can provide to people you want to join your (and perhaps other assigned people’s) a specific circle. Sort of like an auto-accept way for Request Permission to join.. Or really, even keep a list for “queue review” / “circle requests” that you can review whenever you get around to it.Keep the passcode internal – should it be leaked, then you change it and then glare intensely with beady eyes at the person you suspect let the passcode leak..
Nice post. Obviously it is a matter of lowering walls. You do not care if someone not in your circle hears the tune and you never know, that ‘not friend’ may want to become friend.There is a way to do it without need for all the sliding, pushing, backward, forward.
With respect to overly wide circles whether or not you care surely depends on what you are saying? So I agree you don’t care if someone hears a tune but there are plenty of things you might rather people didn’t hear.
Good point. That is a matter of an either/or where you direct sharing specific (contact/group) or wide open. Combination of learned behavior and what platform the sharing is on. Not wanting to sound conceited, but we got that.
This is a very timely post. There have been several interesting discussions on Quora about the same topic.It is extremely hard to assess what the mass market, as opposed to early adopter, reaction will be to circles and we can safely assume that google will continue to evolve the offering, but what is already very clear is that even amongst tech literate early adopters who are positively inclined to google there is a significant lobby who believe that organizing one’s contacts will be just too much work.If we set aside algorithmics for the moment, I think that what this post and others makes clear is that ease of maintenance of groups is critical. This suggests that the UI must enable you to easily change the circle(s) a person is in very quickly and easily pretty much anyplace you see their name. And even in a world in which algorithms provide an increasingly accurate set of suggestions such personal fine tuning will be much eased if you don’t have to ‘go somewhere’ to do it but rather can do it incrementally as you come across contacts in your online activities. Hence you see person ‘X’, note that they are under ‘acquaintances’ despite circumstances having changed and algorithms not having yet suggested a change, and can easily change to say ‘friend.’ I suspect such an ability should be ambient.
You are correct. It is a matter of realizing that the avg user has friends/family/diluge of HS folks/business circles. We are now entering the contacts in circles to the outside whom you may not meet but laugh at their random thoughts/posts/vids/songs.Without doing an over expansive explanation regarding references to ‘biological’ it is a matter of enabling the smoothest interface/interaction crossing the niches where you’re not bound by niche this and/or that. Remember there are some who will be in more than one niche.
I think google’s probably a little gun shy after buzz debacle where they assumed you wanted to share with everyone in your gmail address book. So I don’t think they’re going to rush to automate circles. mis-sharing music is no big deal generally, but more personal stuff could make google look bad again.
Yeah I think you’re right, Jim. Was talking recently with someone who’s helping to lead the project and asked specifically about the issue Fred raises. His response led me to believe they perhaps would’ve been more aggressive in auto-generating circles if not for the backlash they had with Buzz.
Checking katango right now, 10x for the recommendation. I am not impressed by the initial results, but I guess it will learn with time.For me it makes more sense for such a tool to be also available in the browser. Does someone knows about such?
I was going to say that too. Why is it a mobile app?
Hardly any of our family or colleagues or clients have signed up to Google+ yet. Until then, we will not be able to test circles and we just end up blasting/broadcasting into the public. There is no point in segmenting our “broadcast” friends.Maybe in a few years from now, Google+ and gmail are integrated in some way, making circles a perfectly natural concept.
i got my daughters on it. they didn’t take to the social networking stuff but love hangouts
I like the idea of software suggesting circles for you, but there’s another issue you identified:”I did create two groups that I think are particularly valuable; my family and our firm. … Am I going to maintain a list of all the people in all of our portfolio companies? No, I am not going to do that.”The problem is not just that the portfolio company circle would change to often, but rather that you don’t care enough to put it all together. It’s alot of work to maintain and there’s no reason for you to do so.However, just as you feel strongly that you need a circle for your firm, I’m sure each portfolio company has already assembled their company as a circle. What if you could embed their circles, instead of maintaining your own? The people who care about them (i.e. companies) will maintain them and you just have to ‘circle’ them and trust them to do the work.This could also work for small communities (i.e. I follow a ‘NodeJS’ list on Twitter but have no idea who all to follow beyond the 4-5 people I’ve noticed), but it probably wouldn’t work so well for your Music example because it’s such a specialized group.
that’s a great idea
I think that will be possible once Google releases the Business accounts options.
I am just seeing my family (who are not technical aficionados) starting to use. The Circles are going to be a huge plus for most mothers I know (who are the power users on Facebook)… mothers who have been awkwardly using Facebook as a communication and life organization tool..just found what they needed.I thought this was going to be the case & it is proving to be true.
Combine this with my passcode idea (it’s own comment) and create a semi-automatic way of allowing self-organization.
I think that would solve so many problems
Yes, on all counts. Organizing information or groups of people is a drag for most users and should be simplified and auto-aided. The blend of implicit and explicit provides a good balance. What do you mean by ‘created twitter list automatically’? I wasn’t aware you could do that.
a third party service called conversationalist did it for me
A few days late to the conversation but thought I would chime in. I agree in the machine approach to grouping, with a caveat. Social sites need to better utilize specific classifications by non-people entities using the service up front. Take Facebook for example. They treat every page the same, even though most pages have taken the time to classify themselves as a brand, artist, etc. You then take these classifications to help build your groups. Not harping on FB per se, but I believe EdgeRank is doomed. They should allow users to replace the “Top News | Most Frequent” tabbed structure at the top with something like “Most Frequent | Occasional | Brands | Artists …..”. The “Most Frequent” and “Occasional” tabs are organized by your personal interaction with all entities (humans included) on the service. This is based not only on the direct interaction that you’ve had with the entity, but also on the amount of time you have spent consuming the entity’s content (they can fairly easily determine this by observing the amount of time between scrolling in your feed).So I think the answer is a machine approach with a little top-level input from the user as to how they want to see their data. I’m not big into organizing either, but allowing a little housekeeping could go a long way.This goes for Twitter too. I believe they need to allow all entities (people included) to classify themselves on the service. Then lists can automatically be built and turned into tabs on the feed with the option to see By Latest or By Most Frequent. When I want to catch up on things, the last thing I want to do is scroll through the chaff to get to the wheat.
DISCQUS, DISQUS, DISQUSStill the best potential for the ultimate, naturally created, interest based social graph.
Not all my friends would fall into “interest graphs” – not at least a few of my close female friends would seem bizarre if they were algorithmically chosen for me. And yet they are among my closest.I’ve given up on machines doing that kind of bidding for me (to some point)
Smells like an opportunity for a young startup or perhaps Disqus. 😉 Seriously though it will be interesting to see what the G+ api offers us.There is a real danger for machines making those decisions for us. As much as it simplifies it may also filter in a way we don’t want to. I like contrary views from time to time, which may be opposite of an obvious grouping (politics, sports, etc).Regarding the power of circles, I think keeping the number of circles relatively small makes sense. But I do like the idea of having a locals circle, startup/entrepreneur circle and dev circle, in addition to friends, family and acquaintances. This lets me target my stream in a way that wont be filler/nonsense to those outside the circle, and frankly isn’t all that hard to manage once I got the first few hundred out of the way (which absolutely took some time).
Again, I am just seeing my family (who are not technical aficionados and the ones coming from Facebook mostly women) starting to use. The Circles are going to be a huge plus for most mothers I know (who are the power users on Facebook)… mothers who have been awkwardly using Facebook as a communication and life organization tool..just found what they needed.I thought this was going to be the case & in my circles it is overwhelmingly proving to be true.(FYI -The video integration with the Hangouts and the circles is a huge huge game changer for me as well I have to say. We are able to gather much of my family and my nephew who is serving on a boat in the pacific ocean with wifi…load a rented Youtube video and watch a film as a group. and …reinventing how we -as interactive video/multi-plat content creators do post production and dailies)
There’s limits to the ability of technology to mimic the multiple processing that we naturally do in the real world. Of all the social platforms, I find Twitter replicates the ebb and flow of real life relationships; Twitter lists are simple ways to cluster; conversations happen and start very naturally. Now if Twitter could build into is algorithms the kind of functionality you are looking for in terms of close/not close, etc., and if Facebook or Google+ could integrated that into their presentation of our social graph, that would be a cool thing. Oh, wait….maybe that’s what’s Google was thinking with Buzz?
Google circles is the most inviting implementation of visual address book I have seen so far (and that is not a prototype). But, even though I spent at least 30 minutes organizing my contacts in this intuitive interface, I found myself creating a spaghetti incident literally. Mostly because I have trouble defining my friendship proximity, but also because google is really not helping me much with suggestions. And they should and I am sure they will very soon.Nailing circles of friendship has been a hard nut to crack for a very long time, but it seems that google, potentially facebook – via machine learning mechanism can now actually do it for you. Or at least get you half way there. They have the data, they own address books and connections between address book entries and they have the data of “distances” between personas.But there is still one large bit missing that would hugely improve autosuggestions capabilities in visual address books. The biggest communication channel is still phone and address book in the phone. Operators already measure social graph by analyzing phone number, connectivity between them and length of communication. But in most countries they are/will not be allowed to expose or commercially use this kind of data. But Google can do the same via android, and that puts google in a fantastic advantage compared to operators, Facebook and the rest of the address book owners. They have the quality and quantity of data, they are worldwide present and obviously they are not limited by the same rules as institutionalized operators.…. I think the game of who owns address book just resulted in a total winner.I hope they will start using this data, but hell, I also hope they wont abuse it.And btw, even though I like the idea of having a person in multiple “folders”, I also think it brings more confusion than added value. I seriously think it was more of a engineering decision than a usability one.andraz from Toshl
I completely agree, Fred. The “work” component of Google+ is making it seem more like a chore than a “social” network experience for me at this point. I *think* I may love it once I get a bunch of circles established (and more of my friends join), but I’m far from that and not sure if I want to invest the time to get there. And the more finite circles you speak of are already ones I’ve established in Facebook, which makes me wonder why I should duplicate the same circles / groups on Google+….
I wish I could make circles concentric.My AVC circle would be within my tech circle.My Stanford Savoyard circle and my MIT MTG circle would both be within my theater people circle.
There’s actually a whole hypothesis around this(occurrence based data binding) , and mother nature seems to use it too.See:”Among the key findings: Neurons are quite adept at their job. “They can pick out a signal from hundreds of other, similar signals,” said Forger, an associate professor of mathematics in the College of Literature, Science and the Arts and a research assistant professor of computational medicine and bioinformatics at the U-M Medical SchoolNeurons discriminate among signals based on the signals’ “shape,” (how a signal changes over time), and Forger and coauthors found that, contrary to prior belief, a neuron’s preference depends on context. Neurons are often compared to transistors on a computer, which search for and respond to one specific pattern, but it turns out that neurons are more complex than that. They can search for more than one signal at the same time, and their choice of signal depends on what else is competing for their attention.”http://www.sciencedaily.com…
Fred, I feel the same way. G+ ‘circling’ is too big and undefined for me.At my startup Rexly (we’re making iOS/iTunes social), we strive to have our cake and eat it too. We deliver our core value prop, “find music through trusted friends,” by asking users to create a short list of their top 6 musically trusted friends. Like a G+ a circle, this friend sub-group is secret (so it’s not biased). But to make sense of your wider social circles, we’re learning how to infer relative trustworthiness to further personalize what social/musical activities to display. This combination of explicit & implicit trust is powerful both from an algorithmic standpoint (a user-defined shortlist is steroids for collaborative filtering) and a UX standpoint (contextualizing recommendations based on stated and previously unstated preferences).Great post. Long time listener, first time caller…
i hope that you continue calling
I like “oppportunity” with three ‘p’ as a possible brand 🙂
We should have used that in our oppportunity fund name
My problem with circles is slightly different. I like using it to restrict sharing to family, close friends, and co-workers, but past that, it doesn’t seem to do a lot of good.Instead, I’d love to be able to say, “Hey, I write about these five topics on a regular basis: mobile technology, sports (particularly basketball), comics, politics and my life. Pick any or all when you follow me.”My problem with Facebook right now is that I have a lot of people on it who aren’t interested in what I want to talk about most of the time (technology), but are interested in what’s going on in my life. Twitter is the opposite. In neither place have I managed to connect with people sharing other interests (sports, comics, politics).Circles solves the problem of sharing with my family and close friends. It does a good job of going from binary relationships to more complex ones. Which makes it more disappointing that it still treats what I share in a binary fashion–you either get it all or nothing.
I just responded above with the exact pain. Quora might be the closest in the fact that you can follow topics, specific questions, or people. However, the same binary relationship exists if a person follows you. Anything you do/write on the site shows up in their stream.
I’ve often thought Twitter should let followers sort themselves and I think the foundation is already there.If I can direct my tweets to people by starting with @, I should be able to direct my tweets to interests by starting with #.I’m a tech company CEO, college trustee and do some advocacy in adoption and orphan care. Those are three very different audiences and I’d talk more to each but not at the risk of annoying 2/3rds of my followers.It would also be insanely cool to know which reason my Twitter followers chose to follow me.
Separate graphs for separate services is the ideal model for me
Fred,In other words, are you telling your family members “don’t bother to follow me on twitter since all I do is post about my professional life, stick to following me on Facebook”? Doesn’t that then put a large onus on each of us to maintain the plethora of services.
that’s exactly what i do
I don’t like mixing certain compartments of my life. FB is friends and family, Linkdedin/Twitter Professional etc.
That confuses me, but maybe I don’t understand what you’re saying. What feels natural to me is to use each service for a type of content, not a type of audience.I want to blog my thoughts or opinions, tumblog awesome photos or quotes, youtube a video or soundcloud some audio.And then I want to tweet each of those pieces of content to the audience that makes sense.I could set up three Twitter feeds but that just defeats the [email protected]:twitter is one person with multiple interests and I want the 5,000 people following me to self-select into whatever circle of mine they prefer.
Me too. Amen, Fred.I know all these services wants to be the end-all be-all graph to end all graphs.But to me, and to you, and to surely MANY others, I’d rather have a sense of how I am guiding the graph along and want to moderate it and modulate it in terms of getting the most out of a particular service. I want the granularity of having separate graphs in each.The bird’s eye view of a meta-graph composed of many other graphs IS possible with interoperating APIs and user-level authorization. That’s the better vision.
Aaron,I share that desire. Not only knowing reasons when somebody follows but for somebody like myself who has less than 300 followers on twitter, I would love to know the reasons that people unfollow me. Was it because they originally followed me because of a tweet on financial technology (I own one) but when I tweeted five times about the Harry Potter premier on Thursday night they were turned off? Or vice versa?I have always thought it would be great if I could create aspects of myself and associate them with my profile (like a circle people can join). When I published I would publish to one or many of these aspects. I imagine that for most people this would be limited to a number less than 6 aspects (family, professional, set of key interests/hobbies).
And I’m arguing that could be implicit and automatic.#sc Check out the college budget proposal we’ll be voting on tomorrow night.#RiskalyzeYourself We release Beta 3 next Wednesday! Sign up at http://riskalyze.com to get a backstage pass.Twitter could easily scan my tweets for the last three months, figure out that I have two sub-feeds and offer follow buttons on my profile for them.Those who follow me get everything; those who follow a sub-feed get just those tweets in their stream. (Just like the @ symbol for users behaves.)No need to create the sub-feeds…let them form and un-form by virtue of your tweeting activity.
Aaron,So in your scenario though if I am a follower of you and am only interested in “#RiskalyzeYourself” for example, how would I just see that content? It seems to me like these implicit channels need to be explicit in a way that your followers can specifically listen to the one they care about.
Implicit creation of a subfeed that you could explicitly follow.
This may sound strange to say -even though sometimes I would not pay attention, actually, I would feel better knowing more about you – it makes you seem more human
It’s interesting that you say that you’re not interested in maintaining lists of people, but go on to describe various services like Twitter, Tumblr, Facebook, and the lot, each with their own separate list of people you follow. I’m also not one who likes to file & organize lists, but I’ve done just that. It’s just that I maintain one list for Twitter, one for Facebook, one for LinkedIn, one for Tumblr… instead of many people lists within one service, which is what Google+ hopes we’ll do. It may just be easier for us to wrap our heads around the notion of separate lists for separate services.
It’s a good point. You have to think of all this in terms of the broader ecosystem.
it is certainly easier for me to do it that way.
This is precisely the stance I am taking with my product in development – rings of intimacy (sure, call them circles) but leveraging the user’s pre-established networks, platforms and social graphs. However the backbone of the platform is cloud storage for just your self-generated content, targeting families, where your content lives privately & from there you can share discreetly P2P with your immediate & extended families, close friends, or email push notifications to specific individuals, push to FB & Twitter for broader dissemination (+ whatever the app du jour in 5-10-15-50 years). So if Fred took 500 photos on vaca, he will upload and share all 500 with his wife, kids, family and, if appropriate, discreetly share the subset of content that suits others by P2P (close friends) on the platform + email notification for those not, and push to FB for a smaller subset of content but larger audience, then twitter, etc. To me, this is the best solution for the multi-platform, multi-device world we’re living in where the user controls, prioritizes, organizes & owns their content privately in the cloud. Not every piece of self-generated content is *social* – but can be *shareable* – particularly when you look back on a life full of content + decades more content to come.
” I don’t have an iPhone so I have not tried it out.”But you have an iPad right? So why not try it out on that?
good point dave. i don’t think of an ipad for apps. i use it for reading in the browser adn on the kindle app. i need to expand my notion of it
Interest based apps and social networks will eventually do this; there’s no reason to trust one company or product with all of your social info. Take a look at strava.com, they do a great job of this for athletes, and the social network is based on real world activity.
Glad to see you posting on Google plus, not much going on there yet.
You’re absolutely right, Fred. This is exactly what I was referring to me in my TechCrunch piece.The most obvious of these to me is geography. My friends move around. I travel a lot. There’s no way I’m updating everyone’s location as they move around. But this could be done automatically.This has HUGE utility and is a natural segmentation. If I want to see what’s happening near home, click on that group. If I’m traveling to a new city, I could instantly see who my friends are and what’s going on.
that was happening for me on my recent europe trip with foursquareit is an awesome thing to behold
I suggest that direct emphasis on ‘social’ is heavily putting the cart before the horse, entering via the wrong door.There is (1) ‘behavioral data’, (2) ‘social’ data, e.g., from the ‘social graph’, and (3) ‘interest’ data, e.g., from the ‘interest graph’.For a person: What a person wants is to improve their (2) social graph and (3) interest graph.I suggest that it is easier to start with data on (3) interests and use it to estimate data on (2) social than to start with data on (2) and use it to estimate data on (3).So, for a person, the data on their interests is ‘primary’, and estimates of their social graph can follow as one consequence.Then trying to improve a social graph without data on interests, e.g., from behavioral data, is not very promising. And a person’s social graph actually doesn’t say much about their interests. However, some behavioral data can good for estimating a person’s interests.For ad targeting, again the interest data is ‘primary’, and (1) behavioral and (2) social data lead to ad targeting only indirectly via, essentially, trying to estimate interests. Trying to estimate a person’s interest data from their social data is especially noisy and inaccurate.Thus I suggest paying more direct attention to interests and then letting social follow.Disclosure: My work is to let a person improve their ‘interest graph’ directly. Yes, it is clear that there are, then, opportunities to improve their ‘social graph’! It is also clear that using their social graph to improve their interest graph is next to hopeless.
We seem to all be going along with Google Plus, and trying it now, but I’m not a fan of building these new circles either, because the problem is in the management after the initial build (like you implied). I wonder if G+ is trying to do too much. It’s a social content stream, a social network, a conversation place, a jumping point for video chat, photo sharing, a micro-blogging platform, etc…This All-in-one greatness is becoming daunting to manage and sucking-up time. There’s only 24 hrs in the day, and I’ve said this before: We are spending too much time on Social…Key is to find the right balance between Doing Stuff That Matters, and Being Social. I’m rebelling against too much Social & trying to find the right balance. What I know is- there is People and there is Content. I want both equally. And I will argue that one leads to another, bi-directionally.
As long as there’s reason to keep you / the user coming back on a regular basis (to any Google service..), and then as long as doing the more irregular tasks for upkeep and tidying (managing connections, etc) are easy then it’s all good.
A friend of mine said to me: “Esayas, have you checked Google+ for some inspiration?” And I typically say No, I won’t. I spend 2 years of my life to figure out that Design Inspires Design is the Problem of All Problems. Just because Google is a big brand doesn’t mean that they have a better system when dealing with grouping behavior. Kevin Cheng put it right with his statement: “The problem with friend Circles, Groups, Lists, etc., is that our brains don’t have a clear information architecture of our social graph.”A circle is not a life inspired metaphor. It’s a design inspires design metaphor. If the web is an abstraction of our brain, circles can be seen as a restriction of our brain capacity.Instead of looking intto FB, Google or Twitter I just looked into the heart to design our current project and was fascinated that it worked.
I’m mostly the same way (I don’t want to do more work managing contacts — I don’t even do that with my own phone)…but I can say that I’m a big fan of having another solid/big option in the social space. It will serve a certain portion of users very well and it will have a net result of more quality content (and noise) being generated.I’m also with you on the idea that the computer should be doing the work for you…actually a lot of the thinking behind knowabout.it is that we should be able to collect all the random and chaotic data you *might* be interested in from around the internet and then have the computer organize it for you into as many simple and easy to understand/use ways (then go the final mile and deliver those results to you wherever makes the most sense for the user personally). …and of course on a personal-interest level, the more content generating services that are out there, the more people NEED http://knowabout.it ;-)Note: in proofing this message, it sort of sounds like a pitch, but it actually was just intended to say I agree with you so much, that I’m building my own startup based on some of these very ideas/thoughts. 😀
Some personal thoughts:At some point, I would love to track how I know various people that I know – the reason being is that every single time I meet someone at this point, I’m adding a little more data to my “who I know” graph.I feel that because of the fluidity and because of the amounts of obscure overlaps, trying to create circles or maps would get very very interesting.And I don’t think it is a me thing, I think once you hit a point of over 300 people you know, everyone has this issue.I keep thinking what this new software does to the traditional a and b lists
Any digital tool is only as good as you use it and therefore it is best that it is reflective of who you are rather than who you are digitally trying to be. So it’s absolutely cool to be abstract and free, as it is to be highly organized.When I recently came across the Feltron Report, it was an eye popping experience of self-reporting and personal organization – the like I have not come across before – so here in the unique case of Nicholas Felton – the way he organizes the digital experience supports and elevates consequently his love of data organization and infographics, and so fully represents the person he is.Feltron Reporthttp://feltron.com/about.htmlReading this piece simply reinforced for me the ethic of authenticity and being true to one’s own life spirit. There is no digital graph or digital experience that is greater than that, but like all aspects of life, the digital experience is best a journey and a discovery – and certainly (at least for me) not an identity or reason to be. Life is not a disability, machines can help ability and speak to our human capability.I have noticed my own habit of embuing words like Twitter and Disqus with anthromorphic qualities – but it is we who are human and the machine that extends, expands and helps engage us towards new realities and how we embrace the quality of our own given life. Whatever digital connection does for us, it is our own choice of being, and whenever we sit with such freedom, we can extend, expand, engage and embrace technology as machine calculation that supports our human decision. The freedom here IMHO is simply not to confuse the process of calculation with the act of decision. The former is machinery, the latter is purely our own human existence.M.
You want to outsource your friendships to an algorithm?
no, outsource the categorization process
I can see google increasing automated circles and suggestions, but I have to give them props for a clean, simple way to go into manual edit mode, which will always be needed.
One of the machines that will do it for you is a camera, and the way this is displayed is through a digital image. There will be more machines in the future. If you can group your family, take a photo with your family and friends, and categorize the photos as “family” or “friends” etc. then technically they are in this group if they’re in this photo. If you can visualize yourselves together (where you share a common physical existence with these people), then you’re more likely to be real connections at one point in time opposed to virtual relationship or simply email contacts. Algorithms based on email or the people you think are your friends because you follow them on a service such as Twitter may not always real connections. That’s why I truly believe the real social graph is who you can physically be with in real life and have a photographic connection. It’s interesting to think about, and that’s what I’m striving to do with my social network invention is to link people in photos.
had same experience on google +. i have close ties and weak ties. that’s it. the weak ties are a confused mish mash of people i know, acquaintances, pple i am social networked too or just think are the bomb.i’m glad google + took so bloody long to appear. if it hadn’t, i might feel differently. I allowed relationships to emerge even when i was a tad uncomfortable bc connecting to strangers didn’t come naturally. But now it does and the value i’ve gotten from so many of those have been utterly invaluable.
I’m a little late to the party here, but the point that Kevin made in his post that resonated the most with me is once you make all these groups, how are you going to remember who’s in what group? It’s believed that humans can only remember up to 250 people, but potentially I have close to 1000 people I interact with at any given time online. Even after I do the organization, which like you Fred I probably won’t, I still have to remember how I did the organization in the first place. I wonder if it would be better to impose smaller limits on these groups, or on the amount of “friends” one can have on any network, period.
mostly agree with you here…but to me limits are the anti-thesis to the internet…
I hear you, but used in the correct ways it could be potentially very beneficial. I wouldn’t want this, so maybe not the best example, but imagine a 200 follow limit on twitter. If I used it as primary an information resource it would force me to truly think about who I’m following, and where I’m getting my information from.The end result could be that I am getting extremely relevant, specialized, and interesting information…
removed by commenter
I really love Disqus!!! I`m the biggest fan ever!!
Totally with our on the fact circles should be automated as opposed to work. The most valuable and logical use of circles for me has been “Clients” – not only current ones but former clients that you want to maintain a relationship with.Mark
Most people will just have Friends, Acquaintances and Collegues. Power users add more. What’s the problem?
I became frustrated pretty quick with circles too. Couldn’t figure out where to put people. EX: what to do with online relationships that I’ve met in person. Don’t see them everyday, don’t know their family, don’t know where they live, but I do trust them. Are the just acquaintences or friends or anew circle of friends met online?Too confusing and rigid for my fluid and undefined world.Glad I’m not alone.Maybe I’ll create a “confused” circle and add you Fred. Anyone else want in?
I think you are missing the point with circles, Fred.Out going: Are those in your “music circle” the only ones interested in your music musings? Why would you exclude others? Perhaps Bono would have a circle with music collaborators or influencers that he would continually want to publish privately.What are the groups I post to? Close Friends, Family, Business, Public. (3 Circles) Public is not really a circle.In going: Do you really read everyone who you follow on Twitter? If you do :). You have to actively filter to get value from the feed. Way easier using circles.What groups do I filter on?Marketing, Developers, Business Leaders, Start-up, Media Commentators (5 Circles)Was VERY easy to put together. It was not tedious, and the UX made the experience in fact delightful.
G+ should automatically create circles like these LOL (via @pogue ) – http://www.happyplace.com/8…
Like everybody, I’m just experimenting. But it feels to me like people who find Google+ circles too much work are making things too hard for themselves. You can start very simply with circles that work for you to follow and share, and add if you feel you’re missing something. And you don’t need to group the network in general. Instead, when you’re sharing something specific,or interacting with someone in specific, you can double check if the groups are right. For example, if I discover that someone I know from some other context shares some musical interests, then I’ll add them to my music circle. I believe that it’s easier for Fred to use different services to maintain different graphs, but that doesn’t seem simpler to me! Plus facebook is an awkward hodgepodge of overlapping social contexts, for example, cousins who live far away don’t care about a local city council campaign. G+ does a great job of solving those Facebook problems.G+ needs to do more to address the interest graph, but I suspect they’ll be working on it.
The answer kinda lies in your post itself. Family, colleagues and close friends are the only well-defined groups. Every other group is fluid. Algos can be good at managing asymmetric relationships (people whom you want to follow), but not symmetric relationships (making distinction between friends, companions, acquaintances and sub-classifications among those). Would you rather spend more time rearranging or purging circles given that any algo to manage them would be imprecise at best?An example where an algo may work well is if it automatically puts XYZ in your Following circle based on your Sparks or if you read XYZ’s blog/column on Google Reader.
This is an insight that’s similar to the discussion around tagging individuals in photos: few people actually like to go through the hassle of curating items. Manual tagging of photos vs. facial recognition with confirmation of false positives? The latter is far easier for most people. I expect to see much more of this “autonomic curation” for users. In fact, Google’s already started to build it into two of their products: Gmail and Picasa.
Yes, and I’m really surprised no one else seems to have mentioned this here! Gmail auto-suggests email recipients anytime you compose a message, which is a pretty great feature. It’s akin to what Fred’s looking for when he wrote: “If Google+ knows who my music friends are then just suggest “music friends” when I hit the share button and send it on.”
I think circles is a good idea, I just need to find a more efficient/effective way to use them. Not usually one to promote in a response, but thought it was relevant to this post. FavorTraders.com enables users to group connections for a clearly defined purpose – giving and receiving help. Goal is to strengthen community and relationships in the real world.
All of the people who are writing about google+ and how to use it are power users who most likely have a large circle of friends, many of which are probably not even real “friends”. I’d say for myself, and I would guess most casual users of social media, grouping of people really isn’t that difficult. Most everyone falls into a distinct category- family, friend, and acquaintence. If I have something to share, its pretty easy to determine who I’d like to share it with.For you Fred, and for many others that have written about the applications of Google+, there are probably many other undefined groupings because you have an extensive online community and have met many people online- and hopefully google will come up with a good way to help us sort, because as you said, most people will find it tedious and probably won’t take the time to do it. For the average user, I’m guessing it won’t be so difficult to sort things out.
I fussed with those circles for about 45 minutes and then stopped and put all the incoming new encirclers into a circle called “Tech” which is the largest now — and I could have just called them all “Internet People” in one giant circle because just about all of them are unreal, not people I really know, and aren’t my real, true friends, much less relatives.I never used Facebook for private communications like “Here’s our new niece!” and “I can’t make bail for X, can u?” anyway — for that, there’s email! I use Facebook to follow those same Internet People pretty much as I would Twitter or anything.IF I were big and important like you Fred, I’d very muchy worry that those circles like “Our Portfolio” in which you might put comms that you didn’t send to anyone else will ultimately get hacked. Everything is hacked, sooner or later.
particularly if murdoch is involved
No. Not at all. Because Murdoch’s people hacked a dead girl’s phone, some parliamentarians who padded their expense accounts, and some other public figures, but they admitted it was wrong and are making amends AND are facing parliament and the law now. Not so the legions of your pals at Anonymous who you think are innovative and edgy. Not so WikiLeaks. They aren’t apologiing.Murdoch’s people *may* have hacked into 9/11 victims’ phones — we don’t know that yet. All identifiable, and they’ve had the highest officials resign over this matter and vowed to correct; they’ve even closed the tabloid responsible. So no, not “particularly if Murdoch is involved” — as if you even read Murdoch papers, Fred, and as if they could even get your IP that way.WikiLeaks did organize a hacking and publication of 9/11 pager messages and like so many people, you have a double standard on this — indignant about Murdoch, silent about Assange.No, what you have to worry about MUCH MORE is all those little script kiddie felon friends you’ve collected around you of the likes of Aaron Shwartz, the “co-founder” of Reddit. You bless hacking and you condemn copyright law constantly, and thereby create *an enabling environment* for hacking and theft. They will come for you some day, Fred — even EFF and Gawker their darlings and fans get hacked. Then remember this message.Maybe Reddit isn’t a company you fund, but canv.as is, and Chris Pools is actually the same sort of ethics-free guerilla warrior against property and paid content. If you can’t actually find any smoking guns tying him to the awful crimes of his 4chan.org users who have caused $171 million damages in Sony, well, maybe you aren’t looking hard enough (and he certainly isn’t). Even Lawrence Lessis has taken a step away from Swartz.
i’ve never met aaron and have no idea if he’s done something wrong or notebut chris is a very decent young man and far from the ogre you make him outto be
What about locked circles and open circles? You could lock a circle, your family for instance, or your firm, and Google’s algorithms wouldn’t mess around with that circle. However, the open circles Google could suggest or algorithmically add users it thought would fit. It would solve both problems.
Fred – I agree but can’t help but feeling like the organization model of Facebook/Google+ is backwards when it comes to using circles or lists for noise control. Just because I add Fred Wilson to my “VCs” circle doesn’t mean that everything you post is going to be about venture capital.IMHO, the right organization model to control for noise (not sharing) is to organize around “topics” as opposed to groups of individuals. Topics are pretty stable over time and people are pretty used to the hash-taging metaphor thanks to Twitter.This is the model that we are just starting to roll out over at Company.com. You can check out the venture capital topic to get an idea of what I’m talking about: http://www.company.com/topi…
yes, very valuable
I’m pretty sure Google will keep all that +1 data out and available to developers. Rich onboarding data. They’re creating a public gold mine of taste data.
I feel like the overlap is the biggest problem. Especially once you hit a certain number of people you know, or would like to get back in touch with, or something. Our tribes don’t separate into neat groups.Actually, I would love to see a fluid map of all the people any of us know. Just to prove this point
Very true. These are the things you can only refine through really using the platform. But I agree with ShanaC, “our tribes don’t separate into neat groups.”To me, I have no problem with one person being in more than one circle. That just needs to be visualized. I have a few ideas myself on how I would do it and I’m sure Google is watching that and thinking about that actively.I’m okay with circle overlap because even if I have a friend person who might be a close friend, a co-worker, married to my sister, and the drummer of my 80’s rock band (made up circles) all in one individual, the information I would share with each circle differs substantially. The idea is inherently good for this exact reason. What you might say to one group, you wouldn’t say to ALL groups.
Disqus is the best example of implicit connections I’ve experienced online.
They’ve almost got too many directions to go, I wonder which way they’ll dig in and shoot forward in. One issue, their service blocks comments through hosted net connections. At work I have issues reading Disqus, and there’s plenty of high quality comments on technical posts out there.
I think you are right “active and evolving, changing, and best expressed conversationally or through activity”. Within Disqus are my strongest signals, my values, and my beliefs. These are things that don’t find environments on other social platforms to be shared.
Hey thanks for the mention! 😉 To be fair, I built/run http://conversationlist.com but it is really an idea from @whitneymcn:twitter (who btw has TONS of great thoughts on stuff)…that tool is just one of the handful of ‘tools for social data’ that I put together under my http://wow.ly brand of things (and not that anyone asked, but http://knowabout.it is actually just an evolution of many of those tools).Fun tidbit: the http://pu.ly tool that is a part of the wow.ly tool set is something I built from an idea from former USV associate @andrewparker:disqus (though twitter now does email alerts themselves so the service, while still live, is probably on it’s way out).Actually – I spend very little time maintaining most of the wow.ly tools these days as I’m really just focused on improving knowabout.it 😉
No question that’s the big payoff.All social graphs become goldmines of taste data. Google’s eventual monetization of this data will come to eclipse AdWords (in it’s present form) as a revenue stream. I say “in it’s present form” because there is a ramjet effect of Google+’s taste graph waiting to happen in that one of the best forms of monetization on that data will be AdWords targeted against that graph. Therefore, when everything is inevitably in full-effect, AdWords will still be Google’s biggest cash cow but with a major shot in the arm from this rich new treasure trove of taste data.