Don't Give Your Users Shit Work
zachholman.com
zachholman.com
As I see it there are three modes of sharing:
The first is where my post is quite innocent and generic, so I just want to declare it to the world. This goes in public.
The second is where my post is pretty irrelevant to most of my friends, and is really only directed at a portion of them. So to prevent clogging the feeds of the rest of my friends, I submit it to a specific group like "Biking Buddies." Facebook can't learn this or automatically set it up. On the other hand, I rarely care enough to only post to a group instead of public.
The third situation is the opposite of the second: there is a very specific group of people that I don't want to see what I'm about to post. Planning a surprise party or uploading party photos from the night before fall here. In that circumstance I only choose to post to a specific group of close friends. Again, something that Facebook can't deduce from my profile.
You can't get around #3 without doing shit work, except by not posting it to begin with. As a developer you can't avoid this: sometimes manual labor really is the only solution to a problem. Until we invent mind-reading, of course.
For example, you may have an excellent piece of commentary about the history of Unix and the state of X, Y, Z in modern operating system design. Most of your non-technical friends, along with your family, in all likelihood do not care about this and will not find it interesting, so it's just noise to them. If you post it publicly, it's worthwhile to the Internet at large, but annoying to a select group. If you post it so that only the relevant individuals see it, and no one else, then the public potentially lose out.
Is there a simple solution here which I'm just not seeing? Is there a social network that deals with this appropriately? With Facebook, you can "unsubscribe", which means hiding a user's posts from your feed, but that seems like overkill. How do prominent developers deal with this? Do they just make their posts more generic and mainstream, and move the technical discussion elsewhere? Or do they just let their non-technical friends and family Deal With It?
Perhaps one solution is to mark a post with "Family don't need to see this", then skip the post for anyone in Family who views their stream/feed/timeline/whatever, probably with an unobtrusive notice which says "post skipped". But then there's added complexity, and — getting back to Zach's main point — it's more shit work.
A post like that belongs on a blog. I like to think of a blog a place for people to learn about a specific topic. Facebook is for people trying to learn about people.
1. I want to share this with my friends and family, it's ephemeral, will interest them, and expresses more about who I am.
2. I want to share this with my coworkers and colleagues, it is relevant to our professional interests. This category also happens to be the "Public" category for me in a lot of cases, because anything I share professionally, doesn't need to remain private.
3. Anything off topic here. Something that lands in this category (while interesting and relevant to you) is, for the most part, uninteresting to the audiences you've made yourself available to.
My personal "solution" is that I typically share personal stuff exclusively on Facebook (non-public posts) and what is more "technical" (comments or questions on tech, products, etc.) on Twitter. (and HN)
For example, I didn't announce the birth of my daughter on Twitter, because I've known the people who follow me in mostly professional settings. Some of those might be friends too, but if they are, they're probably on Facebook too.
Works well for the most part, except when I'd be interested in the opinion of Facebook friends that are not on Twitter… But it won't work for everyone either.
I realize that it's possible that I am missing out on opportunities to create stronger personal relationships with people on Twitter by not discussing personal matters there, but that's honestly getting too complicated. :)
1 - determining the type of content based on words, size, etc. based on the type of content the system would automatically filter it on the receiver's end. if the receiver has typically ignored or deleted 'history of unix' type stuff from me, the receiver/reading app will file it away, and note to the receiver that something's been received, but it's probably not important. Like 'spam/ham' filtering, but as another dimension. (yam? tasty for some, untasty for others?)
2 - using signals of other people in the ecosystem as more signals in the 'spam/ham/yam' algorithm. If I post something that all my geek friends read/share/comment about, those interactions should count towards an 'interest' score on that piece when it's presented (or not) to my mom, for example.
Which actually works a lot like RSS. You have certain named feeds, e.g. "Unix." Then people looking at your profile can subscribe to that feed if they're interested in what you have to say about Unix.
Two possible solutions:
Channels: You create a few outlets, for example "technical", "personal", "food-i-ate-today". Then I could subscribe to your "technical" channel and ignore the res. Guess this is close to what joebadmo posted.
Tags: Almost the same, but a post can have many tags.
The problems with these "solutions" is: what heppens when you create a new tag/channel? Maybe one should auto-subscribe to all of them, and then ppl could remove those they don't want. Might be shit work though.
joebadmo
I think Facebook already does this. No, they don't setup actual groups, but the ratio of stuff your friends (and pages you've liked) post to what ends up in your feed is probably 10:1. That might be way off, but the point is that FB is filtering the feed based on your relationship with the person, how often they post, the content of the post, etc. At least this is what I've gleaned from reading about this in various places, but I'd love to be corrected or get more detail if anyone has it.
Of course, that's a very small amount of work relative to manually placing friends into circles. But G+ does not (yet) have the same kind of parsed personal/profile information, which would require the same mechanism that FB has (deciding who to reveal what parts of your profile to)...and which, as far as I can tell, is not trivial to implement, or to graft on to the existing Google Account structure.
Of course, Google can ALREADY do this for you. No doubt they have mined enough information about each user, including locations of IP addresses, to fill out most of your boilerplate profile info. It doesn't take the EFF to realize the privacy implications of auto-filling your circles with people who don't realize that Google's algorithm has correctly guessed their location, age, school and workplace and is now implicitly exposing such information.
Google +'s circles are completely up-to-you, for better or for worse, and you hit ambiguous parts more quickly: is that guy a friend or an acquaintance? Should I put him in "Tech", "Ruby", both?
But of course, as you go deeper in either product, you'll soon find the same ambiguities and amount of shit work.
a) Creating a Facebook-like profile system, with more regimented fields and discrete data.
b) Totally disregarding users' privacy by doing it for them.
All of these are technically possible for Google, but they also have their major drawbacks. Where Google has been able to do it without downside - your private list of closest GMail and GChat contacts - it has.
The point here is that for facebook when your contacts change their details, changes to their inclusion in your groups are automatically cascaded.
It's orders of magnitude less shitwork than Google+ where you have to manually curate / update circles.
Circles have great utility for me for two reasons.
1. Like Twitter lists (which I use, thanks Tweetdeck), I want to see information from certain groups of people for different things.
2. I want to disseminate different types of information to different groups of people.
If you don't find either of these use-cases compelling, there is nothing stopping you from ignoring them completely.
I will never ever ever trust an algorithm to get this right, except for the most trivial cases, and if the case is that trivial, I will default to public.
I think ultimately the problem is that these features are trying to replicate offline social context, but only getting halfway there. I've written more about this: http://blog.byjoemoon.com/post/11670022371/intimacy-is-perfo...
I've got to say, this is very much a complaint that doesn't appear to have a good solution. Beyond any sort of simple AI that groups based on last name, school or employer (like Facebook does, I guess), there's not much to go on when it comes to automatically creating Circles.
To me, this is shit work (that the author was talking about).
Yes it's an optional feature and people can not use it but frankly the whole product is optional and people can not use it.
If you're going to put a feature in your product then you should do so in a way that makes it as good as it can possibly be - and that includes reducing the associated shit work.
I use Google+ heavily and I've put everyone in the same circle.
It's a great marketing feature though!
He's absolutely right, though, that a key problem with circles and lists and other shit work is that it's very rarely well integrated into the clients.
(I also want disagree with the claim that "no one wants to do shit work". I believe the entire genre of MMORPGs -- even, dare i say, RTS' -- stand as testament against. They also suggest how high peoples tolerance of shit work is if it is well integrated into a client ;)
> His main point is that adding an assortment of labels, tags, and priorities to your email inbox only serves to give you the illusion of getting work done.
Which I understand, because Mann has a tendency to get into the fiddly bits, and so do I, but what's missing is that these things do potentially have utility. Case in point: I get dozens, sometimes hundreds, of non-spam emails a day. I used to use Apple Mail until it couldn't really handle the volume very well, then I switched to Gmail and learned the keyboard shortcuts. Eventually I got into labels. The UI makes it super easy to apply labels, and I label every important email. This could be considered "shit work", but it provides a solid ROI because it allows me to browse through project summaries, and makes it much less likely for things to slip through the cracks. It's amenable to automation in that I can create rules, but mostly it relies on my ability to tag every single email. It sounds like a lot of work, but once the system is in place it doesn't actually take any time to hit 'l' and autocomplete a label or two.
Meanwhile, Google's attempt to improve productivity without shit work—Priority Inbox—actually provides me negative value. It doesn't matter how good it is because if it's less than perfect I can't trust it, and it can't ever be perfect because countless externalities affect my idea of priorities. In the end, the assigned ratings become more noise that I have to deal with.
So while the point about not letting busy-work make you feel productive is a valid warning, it doesn't follow that if it can't be automated it isn't useful. It's all about ROI. I think the problem with Twitter and Google+ is that they just aren't useful enough to sink that much time into unless there is a direct professional purpose.
I can't believe people are arguing about the right way to do this.
Twenty years ago, it was rare for someone to call everyone they ever knew and scream into the phone how drunk they were. I might have done that only once or twice in my life. (If I called you by mistake and woke you up at 2AM, I apologize.)
But today, if you can't provide a web-based service that not only allows you to do that very thing but protects you from the consequences of it, people will complain.
If it were push communication, you'd be right to complain, but since it is pull communication, other rules of etiquette apply.
Sometime shitwork needs to be done because you can't simplify. I'm doubtful that Facebook's auto-group feature would work for me. Maybe me doing shitwork on Google+ is what work for me, because I value my freedom to control my information online.
In fact, one Google Research paper[1] opens with the line: "Although users of online communication tools rarely categorize their contacts into groups such as "family", coworkers", or "jogging buddies", they nonetheless implicitly cluster contacts, by virtue of their interactions with them, forming implicit groups."
I'm curious what the eleven authors of this paper thought as they saw their Google co-workers developing a system they knew couldn't work.
[1]Suggesting (More) Friends Using the Implicit Social Graph (http://static.googleusercontent.com/external_content/untrust...)
By pushing the privacy aspect of Google+ it allowed Google to differentiate themselves compared to Facebook. That message has persisted.
Users say they care heavily about privacy, but in practice they don't[1]. Circles isn't a bad solution to that, except for the small minority of people who feel the pressure try to build themselves a compete categorisation of everyone they know.
[1] Occasionally people do care - picture sharing is one case where people are somewhat careful. Circles caters to that case quite well.
That anecdote is not really worth much. Especially because those lists are probably very important for those who do use them.
Most people won't care about filtering their social relationships, until they do care. At that point, you want them to have the option.
Many developers obsess over the edge-case of "how do I post secret information that is only shown to the correct list?"
When in fact, normal people just want to post "Going to Aunt Edna's tomorrow!" to their family list, because it's irrelevant to non-family.
It seems Facebook agrees, with their loosey-goosey smart lists.
Seriously. The 1 click it takes to put someone is a circle isn't really "work", and if it saves me awkward calls from my mom because she read a post intended for my drinking buddies, then it was well worth it.
What he isn't praising is when a user compulsively uses a feature which is superfluous. A huge assortment of different tags and labels in email isn't really justified. Sure, you can find a use for it. But most users arguably organize their entire mailbox into these neat categories, and thereafter, half of them are never used again.
The nice thing is, Gmail supports both of these use patterns equally well.
Ideally, that's how every product should treat it's more advanced features. If a user spends the time creating taxonomies they'll never use, well, that's more the user's fault than the software's...
I disagree. I like being able to set filters and granularity. I like having the option to give me the information I want, in the way I want it. I'm prepared to do the extra up-front setup to get the better end experience; I have hundreds of filters set up on my email, for example.
Don't give me forced simplicity; give me the option to tune it to my needs and give it the power to make it actually useful.
Instead, because of the lack of options, we're left with an unusable wall of notifications. Smaller projects get drowned out by larger ones, and it tells me every little detail of what is going on with every single project. I'm pretty sure I've never clicked on a single link from it in the past few years of using GitHub.
As an analogy, Circles is driving users to a brick wall hoping they will climb over to see what is the big deal is. Assuming they care. Assuming they aren't in the middle of something when they get the invite.
Which makes me think about the "Find My Friends" App on iOS. If Apple flipped the "social" switch, they would have creatively acquired a power which no other social network would be able to grasp without huge privacy backlash - knowledge of where you and all of your friends are at any given time. Here's the sell: You already have the app. How does this relate to shit work? Well, you'd be apart of a social network where you, your friends are already members, your latest photos are already there (iCloud), you know where your friends are and what they are doing - and you did very little work.
Automatically coming up with criteria to filter by is a great solution. I'd love if I could send a tweet just to my UK followers or to those who tweet about "Ruby" a lot. This is all easily solved by machines and doesn't require me to do anything by hand.
This made me laugh.
I half-agree with the post. There is an element of "shit work" that actually makes users feel engaged. For instance, my iGoogle homepage has feeds set up with sites I've had to hunt down an RSS for and manually enter. I've had to invest time into rearranging the layout to my priorities.
It's 'shit work' in that it's manual and somewhat trial-and-error, but it leaves me feeling more invested in the product if I'm ultimately more satisfied with the end result.
Disclaimer: I am a PC/ubuntu guy, so I understand that mine may not be the mainstream opinion.
They are a pain to maintain, however.
The author of the article hasn't completely thought this through though. He's saying it's shit work which Facebook automates for you but then he goes on to say relationships are complicated and some people are in overlapping "circles."
He shoots himself in the foot right there. Facebook's auto-populated groups can't figure out the complicated nature of our relationships with people. I have many shades of friends and people who have varied interests even within those shades of friends. It's too hard for an algorithm to be able to deduce this very human aspect of relationships.
Edit: That probably came across more snarky than I intended. The point was that this is easily said, but hard to do right.
Motivation is a big factor as well. G+ the motivation is vague, what benefit do you really gain? I know a bunch of people that jumped on Facebook's lists bc of privacy concerns and desire to limit dissemination of their content to unwanted people. For them their privacy >> a base amount of "shit work"... so author not quite right
Or make it more valuable. If being on specific Twitter lists drove more followers or was perceived to be important then getting people to put you on specific lists would be important.
But it isn't. Now.
In the new GitHub project page (e.g. https://github.com/cocos2d/cocos2d-iphone), it really took me some times to figure out where is the DOWNLOAD button..Please put the download button back to the top right area (next to the watch/fork buttons), and don't give your users shit work... Thank you!
(Btw, I agree what you said in your article!)
1. List Creation
2. List Maintenance
These are two different activities that in different networks require varying attention and utility out of the effort.
I think most people have relatively clear friend/work boundaries, but even then I encountered a few "Hmmm" moments when putting people in circles. I suspect that most people don't actually want to group the people in their social network - it can take a surprising amount of introspection. Time will tell if that's true.
Aye, there's the rub. IT'S NOT FUN.
Disturbs me that on Netflix I've rated six-hundred-and-fifty-four movies - that's a huge amount of "$#!^ work", yet I did it compulsively and enjoyed it. That's the whole thing about "crowd sourcing", "data mining", etc. - persuade people to WANT to do the work, and they will with little complaint. If your users are viewing it as "$#!^ work", you've done it wrong.
If you organize to speed up your "real work" great. If you organize to organize, that's shit work.
I'm not sure what category g+ circles are in though..
People, mostly women in my experience, loved that "shit work" the same way they loved putting on makeup or shopping for uncomfortable clothes.
Some work is enjoyable. People do it for fun. Like gardening, or knitting, or cooking. I believe social networks are kind of like that, for at least some subset of the population. Privacy probably shouldn't be that way, but sadly, I think we all know how scarce people are who actually care about their privacy.
Really?
My point is that people enjoy putting effort into their public persona and appearance, and put up with a lot of crap to perfect it.
Google Circles is an elegant solution to the wrong problem.
Not exploiting that circles are (approximately) equivalence classes is borderline criminal
Our app "Groups" is praised by people who can use it to ... organize their contacts :)
Guess what's next ... a social network.
I also thought his example was contrived. I mean, is there anyone who's so worried about their social networks that they will hem and haw over whether someone is a coworker or drinking buddy or whatever? And if there is someone like that, well thank goodness Google+ supports their neuroses!
The author had an interesting point, but I found his use of "shit work" over and over to be distasteful enough to not want to read his blog again. It is an overloaded and crass term that only served to hurt his otherwise interesting argument.
I don't understand, are the features themselves bad? Who's forcing you to use Twitter lists, or Facebook lists, or even Circles?
In Facebook's case, they may have written the initial lists feature as way to show they're "doing something" about privacy, so simply deploying the feature is a win.
But in Google's case, they've made Circles a central feature of their platform and it's coming up short (so far at least).
The main ones that come to mind are the ability for other people to create relationships, whether personal or work related. This eliminates some of the work when building your profile. Then Facebook implemented the auto-grouping of lists depending on employers, previous interaction, locations, etc. And lastly, they monitor your graph on a daily basis and serve up possible additions to your friend lists which can be added with a simple click.
Google could have done a little of this using Google data such as Gmail lists and all, but they haven't (yet). Personally, I think Facebook's lists and grouping feature is far superior and I'm actually finding out from non-technical friends that they're actually using the friend lists. I suppose this has to do with Facebook automatically pushing a number of default lists to users...
The hard problem is to figure out how to machine categorize this stuff, for sure. But I don't think it's shit work on your users to give them the categorize features. This stuff takes a lot of machine learning and a big system graph that hints at how this stuff should be applied. I think calling it shit work is over simplifying..
And I didn't even begin to address the amount of shit content replicated across all of my social media accounts by the same people.