Social Computing, Phase 1
Problem Space + Requirements
This document is written for the Final Project in H541: Interaction Design Practice — Instructed by Sonali Shah, at IUPUI in Indianapolis, IN.
Other parts of this project:
The final project for H541: Interaction Design Practice focuses on the social computing problem space. To be more specific, “social computing has to do with digital systems that support online social interaction” (Erickson, 2017). That can include the direct communication of data, general distribution of information to a network and facilitation of group collaboration online.
1.1: Problem Domain delves deeper by examining a specific problem in the social computing space. It also includes a brief analysis of existing products and concepts that attempt to solve the problem.
1.2: Target Users summarizes the intended users of the solution. It also provides the results of the user interviews performed, with an analysis to find similarities and differences in responses.
1.3: Purpose and Goals summarizes how the intended design will go about meeting the needs as described by the user research.
1.1: Problem Domain
One major component of the Social Computing space is peer-to-peer communication. This manifests itself in the form of many digital conversation spaces, from phone calls to email to messaging platforms.
Chapter 4 of Interaction Design: Beyond Human-Computer Interaction introduces the three basic rules of conversational analysis:
- Rule 1: The current speaker chooses the next speaker by asking a question, inviting an opinion, or making a request
- Rule 2: Another person decides to start speaking
- Rule 3: The current speaker continues talking
While this pattern of interactions holds true for traditional conversation (i.e. face-to-face) and phone conversation, it does not hold equally true for email and messaging. These platforms introduce a significant communication delay. Whereas with face-to-face conversations and phone calls, a conversation participant can assume that their message has been received and expect an immediate response, email and messaging embody a cycle of message send, message receipt and an unpredictable delay for a response message.
This phenomenon can be further explored via the concept of adjacency pairs (Schegloff and Sacks, 1973). This states that messages are expected to come in groups of two. The first message sets an expectation and the second provides a response. This pattern is present in both “live” conversations (face-to-face, phone) and in “delayed” conversation (email, messaging).
The delay in conversation afforded by email and messaging can be beneficial; it can provide time to gather additional details, compose a response and be careful with wording. In fact, the ability to deflect an immediate response may be why the millennial generation prefers texting and email to phone calls (Alton, 2017).
However, that delay can also introduce anxiety, impatience and other problems. These problems are explored further in the user research below in 1.2: Target Users.
Who does this impact? According to Pew Research Center, 77% of Americans own smartphones, as of January 2017 (Smith, 2017). Furthermore, as of April 2015, 97% of smartphone users used texting and 88% used email (Anderson, 2015). This includes individuals of all ages and demographics.
Additionally, I intend to focus on the communication delay in terms of personal conversational interfaces, as compared to business-oriented ones.
What currently exists to address the problem of the communication delay? Most of the existing solutions for this problem involve some kind of automated acknowledgment that the message has been received.
The paradigm of this is the “Read Receipt,” a term introduced by the iOS messaging app. When enabled, this feature displays as “Read [at time]” in the user interface on the sender’s screen.
This feature can be enabled globally for all conversations, or for individual conversations. This means that if you enable the feature globally, anyone who sends you a message will see a “Read [at time]” when you open that message.
Other popular messaging apps have implemented the same feature using different methods. Another popular messaging app, WhatsApp, uses checkmarks to display when messages have been sent, successfully delivered and read by the recipient.
This feature can be enabled or disabled globally, but cannot be granularly enabled for individual users or conversations.
Facebook Messanger uses yet another approach, by displaying a checkmark for a successfully delivered message, which replaced by the user profile image of anyone who has read the message.
For Messenger, this feature cannot be disabled. It is persistent in all conversations.
A few other messaging platforms worth mentioning are Quip and Hangouts. Quip uses a checkmark system similar to WhatsApp, always displaying in a pattern of: “Seen by [name 1], [name 2].” Hangouts uses floating heads, somewhat comparable to Facebook Messenger, with the key difference being that the images are greyed out when the individual is not actively viewing the conversation screen.
Intercom uses a similar approach to iOS messenger, with the key difference that it does not display the time it was read. Instead, this “Seen” label is only displayed when the responding party has begun to type a response, This is intended to reduce empty time when the message is “seen” but not responded to yet (Donohue, 2015).
In this instance, the feature is enabled and cannot be disabled. But this carries different connotations in a business context compared to a personal one, due to the nature of information being shared.
For the purposes of this project, I would like to focus primarily on messaging platforms. But email shares many similarities in terms of form and function, so analyzing how this problem is solved in the design of email clients is valuable as well.
In all of the previous examples, the user who enables “Read Receipts” is allowing anyone who sends them messages to know when they have been opened. It is permitting extra data about themselves (when they are viewing the message) to be shared with the other party.
Some email clients flip this concept on its head. They allow the email senders to turn on a “Read Receipt,” which will notify them when the recipient opens their message. The control over whether the read receipt is displayed changes hands from the receiver to the sender. Spark mail (for iOS and macOS), is one of many email clients that notably does this.
“Read Receipts” aim to address the problem of communication delay by introducing transparency into the conversation process. While in a face-to-face conversation, participants can gauge nonverbal cues from body language and facial expressions, the same is not available in digital messaging.
The problem trying to be solved is to provide some kind of nonverbal response in the binary of digital messaging. It should be something that requires no effort in composing a message so that the recipient can respond right away. And in that mindset, “Read Receipts” are not the only product feature out there that attempt to solve this.
Another take to solve this problem is Slack’s reaction feature (Noriega, 2015). It allows users to respond to a message without typing one out.
Not only can users select from the standard set of unicode-standard emoji, they can also use emoji customized for their team. This reaches beyond “I read this” and allows for emotional, affirmative/negative and humorous responses, all without writing out a message.
iOS Messages uses a similar feature, which it calls “Tapbacks.” These allow for a set range of responses, including love, like, dislike, laugh, emphasis and question. Similarly, this feature allows for a non-message response (albeit, with much less range in meaning).
Spark, the aforementioned email client for iOS and macOS, has a reaction feature as well. However, because there is such great variety in email clients, all of which can send and receive email to each other, there is less opportunity for built-in reactions. That is why Spark offers “Quick Replies.” This includes a set of emojis, ranging from “like” to “agree” and “thanks.” Spark sends an HTML email in reply including the emoji and word, with “Spark by Readdle” at the bottom. Users are able to customize their own preset messages, including the emoji, word, and accompanying text.
A final solution worth mentioning is “Smart Reply” in Google Inbox and Gmail. This feature provides some suggested short responses to an email, using machine learning. The suggested responses are startling accurate at times and can be used as a response on their own, or as a starting point for a reply message.
While these existing products and features may provide users with value, they do not fully solve the problem of communication delay, as evidenced in the user research outlined below.
1.2: Target Users
In order to understand how conversation delays are currently handled, I comprised a set of questions and interviewed 3 users. I wanted a range in age demographics given the wide group of users that communicate via messaging apps. My first participant was a 20-year-old college student. My second was a 29-year-old experience consultant. My final participant was a 50-year-old school nurse and volunteer coordinator.
The questions follow a basic beginning middle and end format to facilitate a conversational flow. While I asked every question at some point during the interview, I went in an order that was most natural to the conversation, so as to keep it moving smoothly and maintain more authentic responses.
- What messaging platforms do you use to communicate (e.g. iOS Message, WhatsApp, Facebook Messenger, Google Inbox etc.)? What situations do you use each of those (e.g. personal, work, other)?
- Are you familiar with the concept of “Read Receipts”? And if so, do you use those for yourself on any of the messaging platforms mentioned? Why or why not?
- How do you feel when others have “Read Receipts” enabled? Are there any benefits and/or drawbacks when other use them?
- On average, how quickly do you return messages that you receive? If there is a delay in response (i.e. you do not immediately reply when you open the message), what is that caused by?
- On average, how long do you wait for others to return your messages? What is your thought process when waiting for a response? Have you ever faced any challenges when awaiting a reply?
- Tell me about a time that you never responded to a message from someone. What was your reason for not responding? Was it intentional or unintentional?
- If you could change anything about communication delay between yourself and others on messaging platforms, what would it be?
Analysis of Responses
The “Read Receipts” feature is perceived as a double-edged sword. The younger two participants preferred to keep them off when possible.
Often this is due to the fact that the recipient is busy and does not have time to respond immediately. Sometimes, the recipient needs extra time to gather more information, or to get a better understanding of the message in order to craft the right response. “Read Receipts” put on pressure to respond sooner. Otherwise, the thought process for the recipient is, “you read this, and you’re not responding. Why are you ignoring me?”
Another reason expressed by the user research participants for not liking “Read Receipts” was that they wish to avoid responding. One participant talked about an acquaintance who enjoys emphatically discuss politics. This individual is highly opinionated and generally unwilling to consider others’ perspectives. This makes them rather unpleasant to spend time with. And yet, this person messages the respondent often asking to set up a time to “hang out.” The participant avoids responding to these messages, but leaving “Read Receipts” on feels harsh. They described it as akin to posting a note that says, “I read what you said, but I’m choosing to ignore you because I don’t want to talk to you.”
In some contexts, respondents feel okay that “Read Receipts” are a mandatory feature. For example, the youngest respondent explained that Snapchat notifies the sender when you open their message. But this feels okay because he only uses Spanchat with this close friends. They are not bothered if he does not respond after opening as it part of the natural cadence.
But two out of three respondents could see the value in “Read Receipts” (even though one of them preferred not to use them). The oldest respondent said that it seemed useful for the message sender to know that you have read their message, and vice versa.
The other respondent, who could see the value in “Read Receipts,” but preferred not to use them himself said that they seem more practical in a professional context, especially in terms of accountability. Work is constantly busy so it is convenient to let other people know he has seen their message. But after saying this, he back-tracked a bit and said that this benefit does not out weight the negatives. When he is waiting for a response, it makes him uncomfortable to know that someone has read his message but is not responding (for up to a few hours).
One respondent noted that the way that iOS Messages handles read receipts feels harsher than other apps. Because it specifies “Read [at time],” it emphasizes how much time has elapsed. This makes the delay in response between “Read” and new message feel much worse. Compared to Hangouts and Facebook Messenger, the “floating head” profile image still indicates the person has read the message, but in a subtler manner.
They expanded on this by saying that for certain people, “Read Receipts” would not bother him. Most notably, in conversations with his wife—it would be a handy way to let her know that has seen what she has said. If he doesn’t respond, she will not be worried that they are going to “break up” or something. Rather, she will know that he is momentarily busy and will respond soon. Or perhaps, no response is needed.
In contrast to the general dislike of using “Read Receipts,” all three research participants appreciate when others have them turned on. They expressed an increased feeling of control over the situation.
However, in addition to this, many indicated that this can introduce anxiety and uncertainty. If the delay between “Read [at time]” and the next message is too long, it causes them to question, “did I say something wrong?”
One respondent proposed that mostly older people have “Read Receipts” turned on because they do not know they are turned on, or they do not know how to turn them off. He added,
My parents aren’t going to passive-aggressively ignore me on iOS Messages. So that doesn’t really bother me.
This was confirmed by the results from the oldest participant, who was unaware whether their “Read Receipts” were enabled or not. However, they expressed that as a parent of children, it is highly beneficial to be able to see that your child has read your message. Because in that relationship, it is more important that the parent knows the child has read the message than that the child is ignoring them. Because that is what kids do anyway.
In a perfect world, respondents could always reply immediately. But life gets in the way. The youngest participant said:
“If I get a sweet heartfelt monologue from my grandmother about how much she wants to have a call with me, then I’ll take longer to respond. Or sometimes if I’m Snapchatting with a cute girl, I’ll wait a little while so I don’t seem to eager in responding right away.”
Yet another complication that exists within this system is missed messages. All three participants mentioned forgetting to respond to a message and later feeling bad about it. This can happen for several reasons, one of which is when users accidentally open a message without reading it. This causes the message to become mark as “read.” It removes the blue bubble representation in the interface that they have a message to respond to. Messages are also forgotten and missed when the recipient needs to first gather further information or wants to take more time to compose a response, and then forgets to respond completely.
After these detailed interviews, I got a better picture of the existing solutions. Namely, that a perfect solution does not currently exist.
With “Read Receipts,” in each implementation, there is value to be gained in transparency and providing the sender with feedback. But with that, there are numerous drawbacks, including pressure to respond right away, unintentional rudeness and a lack of privacy.
The other solutions presented, namely iOS Messages, are still lacking. On the one hand, they require additional thought in choosing which message to respond with. Additionally, for users with older version of iOS that do not support Tapbacks (iOS 9 and earlier), messages are sent along the lines of “Austin emphasized ‘Let’s get dinner tonight!’” which can be quite confusing and cause a flood of unwanted notifications.
Spark’s Quick Replies are useful but potentially unexpected in the email channel as they are not part of the user interface. This could be potentially confusing for those who are not familiar with this feature or what Spark is.
Google Inbox and Gmail’s Smart Reply feature is helpful as it encourages a response, allowing users to realize that responding can be easier than it seems. It encourages approaching the response with a different perspective. However, it does not solve the communication delay as it still takes time to craft a response. According to one user research participant, the Smart Reply options feel too short to be emails—they seem rude or unprofessional. And at that point, you are still writing up a message, which means that it is just as likely to involve a communication delay.
1.3: Purpose and Goals
[5pts] Provide a focused description of the area that your project will address.
Based on the feedback from the user interviews, there are several goals that must be accomplished for a successful design. The solution should emulate the nonverbal cues of face-to-face interaction, by providing instantaneous feedback without effort from the message recipient. In that vein, it should do the following:
- Provide transparency between the sender and receiver.
- Require little to no effort on the sender and receiver’s end.
- Allow for immediate feedback from the receiver to the sender (and vice versa, if applicable).
- Encourage a quick response time.
- Prevent messages from being forgotten.
- Both sender and receiver should willingly opt-in (on each message send instance or for the entire conversation history).
Digital methods of communications do not inherently contain the same affordances and phone and face-to-face conversations. Rules of conversation which are often dependent on nonverbal cues must operate differently. This seemingly simple problem space is ripe with complications in human psychology.
While the various implementations of “Read Receipts” attempt to provide more transparency into the status of a sent message, issues of consent for privacy and desire for tactful and timely responses limit its use and value. Many users find that “Read Receipts” put an unnecessary amount of pressure on conversations, and therefore tend to avoid using them.
To help solve the problems introduced by conversation delay, a solution must allow for instantaneous feedback that does not overpromise on the recipient’s response time and prevent unintentional misunderstanding.
- Alton, L. (2017, May 11). Phone calls, texts or email? Here’s how millennials prefer to communicate. Retrieved October 28, 2017, from https://www.forbes.com/sites/larryalton/2017/05/11/how-do-millennials-prefer-to-communicate/
- Anderson, M. (2015, April 01). 6 facts about Americans and their smartphones. Retrieved October 29, 2017, from http://www.pewresearch.org/fact-tank/2015/04/01/6-facts-about-americans-and-their-smartphones/
- Donohue, B. (2015). Product principles: read receipts. Retrieved October 23, 2017, from https://blog.intercom.com/product-principles-read-receipts/
- Erickson, T. (2011, June 29). The Encyclopedia of Human-Computer Interaction, 2nd Ed.: 4. Social Computing. Retrieved October 27, 2017, from https://www.interaction-design.org/literature/book/the-encyclopedia-of-human-computer-interaction-2nd-ed/social-computing
- Bullock, G. (2017, May 17). Save time with Smart Reply in Gmail. Retrieved October 28, 2017, from https://www.blog.google/products/gmail/save-time-with-smart-reply-in-gmail/
- Noriega, M. (2015, October 08). Facebook’s new Reaction feature builds popular emoji options into posts. Retrieved October 29, 2017, from https://www.vox.com/2015/10/8/9481469/facebook-reactions-emoji
- Preece, Rogers, & Sharp. (2016). Interaction Design: Beyond Human-Computer Interaction (4th ed.). Hoboken, NJ: John Wiley & Sons, Inc.
- Ritchie, R. (2016, August 01). How to use emoji and Tapbacks in iMessage for iOS 10. Retrieved October 28, 2017, from https://www.imore.com/how-use-emoji-and-tapbacks-imessage-ios-10
- Schegloff & Sacks. (1973). Opening up closings, Semiotica 7, 289–327.
- Smith, A. (2017, January 12). Record shares of Americans now own smartphones, have home broadband. Retrieved October 28, 2017, from http://www.pewresearch.org/fact-tank/2017/01/12/evolution-of-technology/