¯\_(ツ)_/¯

With such small sample sizes, it is difficult to draw many conclusions from this data, and we cannot identify any clear trends in the closed form question responses between the original and updated versions. Question 1′s results appear to be a bit at odds with question 2′s, however several answers to question 6 (”Did the app provide you with any unpleasant surprises or disappointments?”) indicated that the respondent was unclear on what task the mock-up was trying to demonstrate, as well as the context of this being one part of a social network app. One respondent reported that “I wasn't quite sure what task user was trying to accomplish...” while another wrote that “I wasn't totally clear on the broader context of the app. Is this its own social network?” A more thorough introduction to the project and the mock-up may yield different results for question 1.

Reputation and image are key to Slipstream’s success. Specifically, users need to believe that the developers are putting the users first as they design and develop the product, that they can be trusted with the user’s information, and that the user is in control of their experience (e.g. what they share and to whom they share it). Essentially Slipstream needs to be a social network with an Amazon-like reputation for customer obsession.

To this end, I identified revenue streams that make the user the customer. The typical advertising based revenue stream is scorned in favor of micro-transactions, a la freemium games. This will help ensure that the business focuses on the users’ needs, and not those of advertisers. Consequently the key metrics for success focus on growing the user base and reaching an Average Revenue Per User (ARPU) target within one year.

I received the following critique on the original mockup:

The navigation menu is too cluttered.

The navigation menu should contain a list of my recently or frequently viewed streams. All of my posts, whether to my stream, a public stream, or a closed stream, should appear in my home stream.

I should be able to include a note/explanation when requesting an invitation to a closed stream.

Based on this I made the following changes:

Moved the People, Steams, and Tags submenus to the settings menu.

Added a “Recent” section listing your 5 most recently accessed streams to the Navigation menu.

Added the user’s new post in the closed stream to their home stream.

Two points are worth noting. First, while overall the amount of content in the navigation menu did not change, it is now thematically more consistent - it is a quick shortcut to streams the user is likely to want to view. Second, I opted not to implement the invite request note feature, as it seems to me an advanced option that could be available on a closed stream, i.e. a stream admin might require notes or applications from requesters, but I don’t see it as a primary use case. As this mockup is focusing on MVP use cases, I left that out.

Mockup Description

My digital mock up is a Balsamiq project consisting of 28 interactive mocks that walk the user through the following tasks:

Discovery of public and closed streams based on tags they engage with.

Following a public stream and requesting and accepting an invite to a closed stream.

Viewing content within a stream.

Posting to a stream.

Design Principles

Posts are enclosed in a solid black border, utilizing closure to show that all the content within the border relates to the same post. The same format is repeated for each post on the stream, using similarity to convey that these are all items of the same type (namely, posts). Similarly, list and menu items are separated by black lines, but follow the same general design of icon+label. However, items representing streams use different icons to indicate the type of the stream (user avatars for users’ personal streams, a dotted line for public and closed streams, and a hash sign for tag streams). Additionally, menus are broken up into different sections separated by a bit of whitespace, using proximity to relate each sections’ items together. The mockup also uses contrast to indicate when the user has unread notifications by changing the notification icon to red, and to indicate when the user is on the create post view or notifications view by changing the appropriate icon to white.

Feedback

Feedback from the user was that they liked how easy it was to like and reply (not implemented in mock) to posts, and how visible notifications were. They also liked the overall invitation request/receive/accept flow for joining a closed stream. They were however confused by not seeing their new post on the screen after posting it.

Based on feedback from testers, I made the following revisions:

I relabeled the “GROUPS” section of the Streams menu to “STREAMS”, to eliminate inconsistent terminology. (a/b)

I replaced the “hamburger” icon with a gear icon to better represent the settings menu. (b)

On the New Post form, I made the destination stream name a button that when tapped displays the Select Destination Stream view. This allows users to post to the desired stream no matter which stream they are viewing when the tap the New Post button. (a)

I changed the language on the Request Invite dialog so that it explains plainly that the reason the user needs to request an invite is that only members can view the stream’s contents. It no longer relies on the user correctly interpreting the meaning of the ambiguous word “private.” (b)

The subject was a caucasian woman between the ages of 30 and 35 with a college degree, who is active on Tumblr and Twitter, and has a long history of participation in online forums.

As in test #1, the subject found the interface intuitive and unsurprising. Unlike the first subject, this subject intuited why they were being recommended the stream. Three issues were identified during this test:

The term “private stream” is confusing, and sounds like it is someone’s private conversation.

The heading of the “Streams” menu should indicate the currently selected stream.

The subject for this test was a caucasian woman between the ages of 25 and 30 who holds a graduate level degree and is active on Tumblr but not other social media platforms.

Overall she found the interface intuitive, especially “for a tumblr user.” Most of her questions related to items that are not implemented in the prototype (e.g. web version, account creation). The test did however reveal three defects:

The words “community”, “group”, and “stream” were used to mean the same thing, resulting in confusion.

The “hamburger” icon commonly used in mobile apps for a settings/options/etc. menu was unfamiliar to the subject.

Subject

The interview subject is a caucasian woman in her 20s, with a graduate degree, who teaches science in a public high school. She identifies as a member of the LGBTQ community, and prefers not to share personal information online.

Feedback

Big plus for discoverability, but possibly big minus for privacy

e.g. a closeted person would not want this recommendation popping up out of nowhere

Recommendation should be opt-in, or appear contextually when viewing related content

Closed communities need to be able to control what tags they show up under

Subject

The interview subject is a caucasian American woman in her 30s, college educated, and working as a project manager for a major corporation. She identifies as a member of the LGBTQ community, and has extensive experience moderating online communities.

Feedback

Overall positive reaction

Facebook offers this, but has the “real name” policy; G+ offers this, but no one uses G+.

The putting together a how-why graph for the task and user stories for my solutions revealed that said solutions were...not. The first idea, centered around closed, moderated groups, provides the necessary security/safety, but not the discoverability aspect. This jumped out at me when comparing the user stories between this idea and the others.

Brainstorming Session

Top Three Ideas

Use machine learning to determine content warnings and tags for posts, both in general, and from a given viewer’s perspective.

Use weighted voting schemes for user-driven flagging, reporting, voting, etc. on content, to depower the pile-on effect; use factors like amount of account activity, similarity to accounts that have already voted accordingly, etc.

The subject is a caucasian woman between the ages 25 and 30, who works as a teacher. She uses social media in a fairly specific, limited context.

Interview Summary

Tell me a bit about how you use social media: which sites/apps do you prefer, how and when do you use them? I mostly just use Tumblr, for participation in an online roleplaying (RP) community. I use it daily, about 30-60 minutes at a time. Most of my content is RP-related, but I also sometimes share other amusing/interesting posts I come across.

Who is your audience? Mostly fellow members of the RP community. Some random people follow my blog, but I don’t think of them as my audience at all. I also sometimes share things for specific people to see, tagging them in the post.

Do you use the tumblr messenger system at all? Not often, most of the RP community doesn’t use it; we tend to communicate via the Ask feature. I don’t think to use it for sharing posts with specific people, and I don’t trust that they will notice the message if I do use it.

Does the public nature of sharing a post bother you, if you’re really trying to target a single person? Not really; if I want to share something privately I just use email. I don’t share things that are associated with me personally, and try to keep my personal identity separate from my blog. I rarely try to share content with a defined group, so email works fine when I do want to share with a limited audience. That may in part be due to the limited capabilities of the platforms I use (i.e. Tumblr)

How do you decide if you want to share something personal? I ask a few questions: is this something people will find interesting, will I get something (e.g. comfort) out of posting it, is it something I want associated with my blog, should I check with others before sharing it?

Observations

Anonymity is very important.

Having a feature isn’t enough; users need to have faith in its functionality and usability.

Just because someone can see something, it doesn’t mean they’re a part of the intended audience.

Not all communities need strong walls around them.

People’s usage patterns are to some extent a factor of the tools available to them.

The subject is a caucasian woman between the age of 30 and 35, and works as a product manager. She is a long time participant in many different fan communities.

Interview Summary

Tell me a bit about how you use social media: which sites/apps do you prefer, how and when do you use them? For personal life: Tumblr daily, Twitter weekly, Instagram and Facebook rarely (special occasions or when family asks). For work: Workplace by Facebook, and Skype and Slack daily.

What kind of content do you share on Tumblr? Who is your audience? I have different blogs based around interests, each with its own set of followers. Mostly posting/sharing gifsets, videos, and short essays. I also consider content rating when deciding which audience to share content with; I won’t post adult content on a blog with lots of followers.

Are there times where you wished for greater control over who you were sharing content with? 24/7. Lots of social media conflict comes from people being exposed to content they don’t want to see or cannot handle, or don’t understand the context of. Tagging and content warnings help, but are subjective. There aren’t good ways to defines spaces with clear rules and expectations.

What has helped you create and maintain that sort of space in the past? Clear rules, trained moderators with the power to enforce those rules, being open to dialog about rules, their enforcement, and the community in general, and participants understanding that they have some personal responsibility for what they choose to expose themselves to.

If you were to be running a community like this today, what tools would you want to have at your disposal?

Threaded discussions so context is not lost (contrast with Tumblr’s endless conversation forking with reblogs)

Community management tools:

A clear place to post rules and FAQ

Ability to screen out minors

Support for multiple moderators

Ability to freeze threads

Ability to suspend and ban users, and track rule violations by users

Site/app policies:

Clear rules against hate speech

Clear policies on personal information privacy, indicating if and when the policy may be suspended (e.g. for legal or safety reasons)

Content screening for “propaganda” / fake news

No real name policy

Audience control:

Control over who can see my posts, contact me, share my posts

Controls on notifications, e.g. ability to mute posts

Exposure control:

Ability to block posts, users

Ability to filter out content by keyword

“A lot of what Facebook has without the terrifying privacy concerns”

Observations

Many of the desired features are available in different products, but not packaged together.

A product’s policies make up a critical component of its user experience.

Primary concerns are: control what I see, control who sees my content, protect users from harassment.

Meta conversation about communities is important.

Social networks like Twitter and Tumblr have nearly completely supplanted the old topic-based message boards and group oriented journal sites of the early 2000s. The open, public nature of the former makes content much more discoverable, but that discoverability comes with a cost. Users have largely lost any ability to police their own space online, with Facebook being the only major social network that supports moderated communities.

This tradeoff particularly affects members of vulnerable, oppressed groups who have largely benefited from online communities as a way to unite and support individuals who might not otherwise be able to connect with members of their same group, but now find themselves unable to defend their online space from the same sort of harassment they face in the real world.

I would like to explore user experiences for a social network that supports private and semi-private moderated spaces as well as public, unmoderated space, while also enabling content discovery and community participation.

User group: Users of social media, particularly members of oppressed groups, e.g. the LGBTQ community, women (particularly in STEM, gaming, or “geek” spaces), those seeking support for disabilities, illness, or abuse, etc.)

Tasks to support:

Publishing and sharing of content

Management of who has access to that content

Creation and moderation of private or semi-private spaces

Content discovery

Current solutions:

Existing social networks (Facebook, Twitter, Tumblr, LinkedIn, etc.)

Forum software (phpBB)

Journal sites (LiveJournal, Dreamwidth).

Challenges: 

So this heretofore unused blog is getting appropriated for my HCI course, and part of the first assignment is to introduce ourselves via blog post. Of course, this being tumblr, I feel obligated to do so in the form of a meme...

Name: Annabelle Backman

Preferred pronouns: she/her

Time right now: 8:47 PM

Job: I’m a software engineer at Amazon, and the tech lead for Login with Amazon.

Time spent at your current job: 4+ years, about 7 in other companies before that.

Areas of expertise: Internet identity, authentication, and authorization; web app development; backend service development.

HCI experience: I’ve built a lot of bad user experiences.

HCI areas of interest: Designing intuitive interfaces, accessibility, and socially conscious design, i.e. designing for a diverse, global audience.

What you’re wearing right now: Gray sweater, blue jeans, and really comfy socks.

Place you were born: Chicago, Illinois (well, the suburbs)

Last movie you watched: Rogue One

Astrological sign: ♓(That’s pisces in case your browser isn’t speaking unicode)

How much you think it matters: Not at all, but horoscopes are fun to look at.

Hobbies: making elaborate costumes for comic cons, making bad puns, spending too much time on this website right here, being bad at Overwatch on XBox, and adding links to answers in memes.

Favorite TV show from the 90s: Buffy the Vampire Slayer

DND alignment: Lawful Good

Pets: Two 3 month old kittens named Willow and Elphaba (because of course a pair of pagan lesbians named their cats after fictional queer witches).

City/neighborhood you live in: Phinney Ridge in Seattle.

Last thing you ate: Gluten and dairy free pizza and mozzarella sticks from Razzi’s in Greenwood.

How you chose your blog name: It’s a pun of sorts (see above) based on my name.