How Google Can Really Innovate / The Biggest Feature Missing in G+ | Hacker News Reader

How Google Can Really Innovate / The Biggest Feature Missing in G+

plus.google.com

63 pointschristophe97123 comments

Google+'s entire thesis here is based upon introducing another level of abstraction, Circles, into social networking. It's risky. Users generally don't want to learn new abstractions unless they provide a massive benefit or can be mapped onto something they already understand.

Will this complexity be something users embrace? Facebook clearly thinks the answer is "No." (And probably has evidence to prove it.) The adoption of G+ so far is by techie types who love new abstractions and want absolute control, and is no indication of how the world at large is going to react to it.

Crossing the chasm for this product is going to be particularly tough. The population of early adopters (techy types) happens to overlap precisely with the population that will not be turned off by the complexity introduced by Circles. This is a deadly combination, and will likely put up a wall through which the adoption of Google+ will not pass.

This post suggests yet another abstraction, tags, on top of everything else. It's a common reaction: engineer sees a flaw in the degree of control in a design, and introduces a new level of indirection to add the necessary control. Of course, this translates into less adoption, since this makes the product more complex and less approachable.

The Linux GUI is an example of this trap being fallen into over and over again for decades. (It's come a long way, of course, and a large reason for that has been the ability for people to learn how to say "No.")

Google+ isn't just 'techie types'. All of my friends (the ones I care about) have actually picked it up and we're all hooked. My family has joined as well and love hangouts.

Nobody I've talked to thinks circles are complex (most of my friends are non-technical). The abstraction you refer to is not foreign at all -- it is modeled after real-life social circles.

Users generally don't want to learn new abstractions unless they provide a massive benefit or can be mapped onto something they already understand.

With Circles, Google is touching right at heart of human social interaction: You show different sides of your personality to different groups of people. With your parents, you talk incessantly about the grandkids. With your buddies, you talk about old times and shared hobbies. With your colleagues, you talk about your industry. You occasionally mix things up a tiny bit, but you don't go overboard.

I'm convinced that 99% of the human race gets this. (The other 1% are intolerable bores.)

Now, I'm not convinced Google has nailed the UI for Circles. For example, if you want to know who can read your comments on somebody else's post, you need to click on little blue "Limited" link, and try to make out the profile photos. In real life, you would just look around the room.

So there's more work to be done. But I suspect that "social circles" are something that users already understand, and which they'll value.

Right, this is a fair point, and what I think G+ is banking on. The abstraction and additional complexity is clearly there. On the one hand, the modelling part of the complexity is a wash, since as you state humans generally already understand separating their social circles. It's the application of the model that people will find confusing, and will require a learning curve. It might turn out that just getting through the modelling phase successfully will be enough for enough people to sign up, and the complexity of learning how to apply the model effectively can be optional and left to the user to discover.

Circles are great, however, the stream only does an okay job of using that information. All it takes in one hyper-poster or hyper-thread to overwhelm your stream, despite any ordering based on circle priority.

Actually, this seems like a pretty good idea.

I presume that if G+ really is a collection of smaller projects, then this functionality would sit neatly inside the stream area.

It's also one of those ideas that once you hear, seems obvious - so wouldn't be too surprised if they're already working on this. It'd only really be useful if/when there is a large amount of newsfeed traffic on there, similar to how busy facebook is for most of us now.

I hope they're working on it, but like I said, I'm mostly worried about the politics of it.

Also, they might reject that idea just on the basis that their interface won't be as simple with this...

I think the tags idea might be a bit too complex, interface-wise.

I'd prefer an option next to every post in my stream called "fewer posts like this (from this person)". Given enough feedback from me, G+ would automatically start filtering out items for me. Since Gmail accomplishes a similar feat already, this doesn't seem impossible.

However, more practically, I think they could probably just solve my problem by allowing me to mute an entire contact or an entire person. Currently you can "block" individuals, but that's not really what I want - I don't want to block them from seeing me, I just don't want to see their stuff in my main stream.

that would be great, especially with multiple levels of visibility: shown in full length like now, just a one-line note for maybe-relevant and outright hidden for the really boring/annoying stuff.

This seems useful. Hopefully they don't take too long coming out with the Google Plus API, I predict there will be a lot of thirdparty addons/interfaces like we saw with Twitter. I'd like to be able to use circles as venn diagrams and perform more sophisticated unions/intersections/differences/joins on my circles as well as generate them on the fly for one-use purposes.

Not exactly an innovation, but I would be happy if they let me merge my Buzz contacts. Just create a "Buzz" circle, move everybody there and kill Buzz.

Great idea, distinction not on peoples but on the content for your feed. Automagicaly.

I think twitter hashtags are the most close to this, but they lack the automation you suggest.

I hope developers would be able to make bots for g+, like for Wawe.

It is interesting how twitter has de-emphasized the obvious benefit to people creating custom streams through #tags

Interesting idea but to me it's definitely not the biggest missing feature.

What Google+ needs is reddit or HN-style commenting, maybe not with downvotes (for whatever reason they don't already have them, even though they're on YouTube comments) but at least parent-child style commenting. It's been shown all over the web that it's impossible to follow threads of comments when they're all treated the same, Facebook-style, and Google employees should definitely know better as many of them use these sites. I suggested it in the "Send Feedback" area and really hope it gets implemented as it would be a game-changer.

Here's my solution to this problem, which I tried implementing when I designed my own (now-defunct) social networking site:

http://coding.pressbin.com/51/An-algorithm-for-self-enforcin...

Personally, I would like to see something that filters out languages I don't understand. Too often on Facebook my feed is filled up with posts in varying languages that I simply don't know and I would prefer them not being there.

I've suggested a similar feature through their Feedback link, although the other way around: G+ should detect your contacts' default languages (via GMail, for example), and then automatically create circles for different languages (you could amend them manually where you have bilingual people). When you post, G+ should then detect the language (Translate can do that!) and suggest the right circle. To work properly though, you'd also need an "intersect" feature, so that you could share something with "people in English-speaking circle AND people in Friends circle".

Edit: in any case, considering their experience with Orkut (which was overrun by Brazilians and Indians using their own idioms), I bet Google is already working on language filters of some sort.

Why not automate their translation? That's the future of a flat world =)

In reality - what we are saying is that structuring data is useful to saying who we are and interacting with the right audiences. It amazes me how unstructured the web continues to be. It's flexible, but is is valuable.

It is a fundamental paradigm shift - saying that the web should be structured around people. I like it. (and working on it ;))

Rather than tags I'd just like to see custom streams. Just like you select circles and individuals to share with apply the same logic to building custom streams. Then there is not yet another things to learn, interface remains simple and you could filter your stream any way you want.

Automated content tagging of any kind contradicts to core idea of social communication. It's actually step back to AltaVista times.

What G+ needs is one simple checkbox - exclude circle/person from common Stream. That's it.

PS. And #hashtags -- miss them badly in G+.

It also badly needs prettier URLs.

Great minds think alike. I was saying exactly the same thing: http://www.rakkhis.com/2011/07/google-plus-because-i-hate-be...

Reply on news.ycombinator.com