Cat Herder

Project Summary

Cat Herder is a project about improving people’s ability to find and organize their text messages. I began it in March 2021 as part of Google’s UX design certification course.

The inspiration for it arose from my own frustrations with easily finding and organizing texts in Apple’s Messages app.

Figuring I couldn’t be alone, I did some primary research in the form of user interviews with four people. I also did some light secondary research into the API for the iOS Messages app.

The interviews confirmed that organizing and finding texts is a terrible experience. Some people didn’t mind it. Other didn’t seem to mind it until they spoke in detail with me about the solution I was proposing. Then they realized how poor the organize-find experience was and saw how improving it might make their use of Messages more efficient and enjoyable.


Assumptions

Every product assumes certain types of users and behaviors. While they may seem obvious or apparent immediately or in the course of developing a project, I find it helpful to list them in their own section as simply as possible. The Google UX course didn’t overtly suggest I do this. It is more something I find useful, particularly if I want to clarify how assumptions relate to inclusivity and selecting people for user interviews.

Without further ado, here are the assumptions I’m making about Cat Herder users:

  • They use text messaging.
  • The use the most recent Messages app on one or more Apple devices.
  • They exchange personal, professional, or both types of information that they value over texts with other people.
  • They need to find and use this information in texts.
  • They find it frustrating or subpar to find and use this information once it gets buried in new text messages.
  • They dislike needing to manually move information in texts to another app or program in order to keep track of it.
  • Biases alert: They are not vision-impaired or hearing-impaired. They interact with their devices using their hands. They can afford Apple’s products. They have a reliable data connection. They have the time to exchange information over texts. They can read and write. You can keep going down this list for a long time, discovering that every assumption you make has other assumptions buried within it.

Secondary Research and Competitor Audit

I had a few goals with secondary research and a competitor audit (view a PDF version of my audit spreadsheet). Mainly, I wanted a brief overview of what capabilities existed on different apps to organize text messages. I also wanted an idea of how many texts—valued communication with people important to us—are essentially vanishing due to zero or poor ability to organize them.

It’s hard to identify any trend around the very specific issue of organizing individual texts. I can make a few guesses, though, about the broader information ecosystem that might influence the development of Cat Herder. For one, search keeps getting better. At some point, human beings should not have to tag or manually organize anything. A smart search would simply handle this for us, on the fly, and be able to organize information in any way that was needed to meet the demands of any project, moment, or searcher. The idea that we have to spend precious minutes putting labels on things that a simple machine could do … is crazy! Another major trend: the constant fragmentation and centralization of communication channels. Anyone communicating in email, Slack, text messaging, social media platforms, and anywhere else knows there is a huge expenditure of energy and time simply jumping around from one variation of an “inbox” to another. There should be one app to check (and organize)! That is why I included Texts in my report, even though it is still in beta.

Let’s explore just how much texting is going on. Intuitively, I know everyone with a phone is texting, but I wanted to get an idea of how many texts in aggregate for the United States. I want to communicate with dramatic numbers how important this technology is, thus highlighting how crazy it is that we treat information that happens to pass over texts as disposable and forgettable. Luckily, a company called Text Request delivered these facts: On average, American adults sent and received 32 texts per day, totaling 18 billion texts every day, 541 billion texts every month, and 6.5 trillion texts every year.

Can you imagine if the United States couldn’t find 6.5 trillion of anything? What if the government said it had $1 trillion, but it was an old pile of money so it was hidden away, but it wasn’t sure where, and it had no great way to find it, and it didn’t even put it into a vault with a label on it.

We are throwing away fortunes in communication value every day, accepting that something is lost simply because it is older than a few moments. That may reflect the normalized American cultural bias against anything that is not new or young, but normal doesn’t mean okay.

Onto the competitor audit: I know there are problems with the options for organizing and finding texts (see this discussion on Apple’s communities site) that are anything but the most recent ones in a conversation. I knew there would be some solutions to this problem, but I didn’t know how usable, equitable, enjoyable, and useful they would be. What would my competitive analysis reveal?

The closest service I found was Keepster, but it was complex to use and aimed to get people to buy printed book versions of their texts. It also worked off manual and local backups of phone data, and your phone seems to archive messages going back only so far (I don’t understand how the software decides dates of texts to include). This was also very slow, though I was able, after some non-intuitive experimenting, to copy an entire archive of 1,416 messages with my mother (though, again, they cut off at a certain date for unknown reasons) and paste it into a Google Doc. Keepster focuses on the emotional importance of text messages, trying to get users to recognize this enough to compel them to want to create a book to remember them. It is consistent if unconvincing communication. They do a very good job of working through a lot of constraints around Apple’s services to create a book version of texts that people can preview and buy.

Another tool, Groups, required an installation process so complex I immediately dismissed it. I’m not serving the same audience of people who are willing to jailbreak their phones and install software in non-standard ways. The usefulness of the app also seemed limited — you can create groups for different people, but that’s about it. Another app, Group Text Free, provides a similar service, but through Apple’s normal app store process. It had poor reviews, which noted that a group could only accept up to five people. This app makes numerous assumptions about users, and so it isn’t so much selling itself as simply letting people know it exists, if they were already interested.

I didn’t find Telegram to be a competitor because it was structured to be a whole separate service that people need to sign up for, rather than a tool that works within the existing Messages app. However, it had a feature that allowed people to save specific texts into a folder. It was rudimentary but worth noting, since saving individual messages is a core feature of Cat Herder. Here is a YouTube video demonstrating the “save” (or “bookmarking”) feature. It also markets itself based on its security features, not organizational ones, so it is not a direct competitor in that sense.

The app Texts is a bit of a wild card I threw in very late in the research process because it’s an app in development. It centralizes messages from a wide range of sources, including Apple Messages. It offers no organizational tools beyond searching. Yet, given that this breadth of covered networks makes a pretty compelling case to get this app, it’s easy to imagine the developers adding refined search and organizational tools for individual texts, and they would be in complete control of the interface. That would be unlike the Cat Herder approach, which entails working within the constraints of the Messages app and however Apple has structured its API. Texts sounds like an app I would actually use, because I would not lose access to my Messages communication and I would gain an easier way to communicate on platforms like Twitter, Instagram, LinkedIn, and Facebook. I could start organizing memories wherever they occur with an app like this—if the developer realizes this unmet need within text messaging interfaces, no matter where a user encounters them.

Finally, I need to consider the Messages app itself to be a competitor. Apple has vast resources and presumably will continue to improve the app over time. Poor tools for organizing, finding, and using texts in different ways strike me as an obvious problem in need of a solution. Apple will fix it, but I have no idea when or how. The biggest barriers I imagine for implementation for the company comes down to a few things: bureaucratic challenges of getting many different teams behind the changes; analyzing how it affects their business model; keeping a pleasing UX within the app while adding features. It also might be easier for Apple to let independent developers do experiments, so to avoid blame for mishaps and put the burden of testing on someone else.

To explore my findings, check out the quick slideshow, an analysis summarized into a table, and another table looking at the product features that seem obvious to pursue. Essentially, the marketplace needs a simple and fast way to organize individual messages within the Apple Messages app.

Matrix of competitors
Click on image to enlarge in new tab.
Low-Hanging FruitGapsOpportunities
Make an app working within Messages.None exist.Well! Make it.
Provide easy and customizable ways to save texts.Nothing but a “starring” function exist.Organize as many ways as users specify.
Functionality allows to user to organize at the level of individual texts, not just conversations or other texters.None exist, except for “saving” in the closed ecosystem app of Telegram.Enhance a single message within the Messages app, which everyone uses already.

At the risk of engaging in confirmation bias (when you pay attention to information that supports conclusions you’ve already made), I would say we have a significant problem with text messaging that still needs a solution.


User Interviews and Empathy Maps

For this primary research, I needed to weed out anyone who did not use Apple’s Messages app to text, so I narrowed down my interviewee possibilities by asking one simple question: Had they used Apple Messages multiple times in the past week?

From there, I had a few other considerations. First, I wanted users whom I thought would have different experiences with texting. I aimed for a range of ages, and ended up finding people as young as 19 and as old as 76.

Secondly, I wanted different genders and racial backgrounds. I worked with two women and two men. One of the two women described herself as a woman of color and was 40 years old. The other woman (76 years old) and two men (19, 44) described themselves as white.

Their socioeconomic background ranged from middle class to middle-upper class. One of them was unemployed, one worked for a large company, another was self-employed, and the fourth was retired. I would have liked to interview more people from more varied backgrounds (perhaps even using services such as usertesting.com and userinterviews.com). However, I also wanted to complete this exercise in a reasonable amount of time and not spend too many hours on any one phase, as my larger goal is to understand the entire UX design process and how all of its pieces fit together.

I will not post the transcripts of the interviews (email me if you’d like to see them), but you can find the empathy map for each of the four people below. For other UX professionals’ benefit, I’ll note that I used the Rev app to record my calls and Descript to transcribe the audio (its AI is fast, cheap, and accurate, and the editing tool is incredible). I changed the names of my interviewees out of respect for their privacy.


User 1 Empathy Map: Ted, a 19-year-old white male, former barista, and student living in Colorado

empathy map for Ted
Click on the image to enlarge in a new tab.

User 2 Empathy Map: Bob, a 44-year-old white male and web developer living in Philadelphia

empathy map of Bob v2
Click on the image to enlarge in a new tab.

User 3 Empathy Map: Debra, a 41-year-old woman of color and product designer in New York City

empathy map Debra
Click on the image to enlarge in a new tab.

User 4 Empathy Map: Paula, a 76-year-old white and retired woman living in New Mexico

empathy map for Paula
Click on the image to enlarge in a new tab.

Personas

It made more sense to design personas around user goals (goal-directed personas) rather than more specific demographic information. Why? Because that’s what my research told me. Everyone, coming from all kinds of backgrounds and possessing a variety of identities, used texts for every reason you can imagine. But it wasn’t difficult to see how this texting activity could be categorized in two broad ways: coordinative and emotional.

By coordinative, I mean the information in texts revolved around appointments and making them. Think phone numbers, addresses, meeting times and dates. I called this persona Debra Coordination.

By emotional, I mean the information centered on enjoying, sharing, and exploring feelings — so funny jokes, nice reminisces with an old friend, or even mundane texts that seem suddenly important because someone in your life has died (I had this experience with the passing of my mom). I called this persona Joe Reminiscing.


Debra Coordination persona (coordination-oriented):

As a busy person trying to make important meetings, I want to quickly find and act upon appointment information buried in texts so that I can have more control over the information I need to coordinate connecting with others.

persona for Debra Coordination
Click image to enlarge in new tab.

Joe Reminiscing persona (emotion oriented):

As a reminiscing person who wants to feel connected to others, I want to easily find all types of information buried in texts so that I can enjoy, use, and re-share it.

persona for Joe Reminiscing
Click to enlarge image in new tab.

Problem Statements, Pain Points, and Hypotheses

I found problem statements, as taught by Google, to be very similar to user stories. They use the same structure, but with a slightly different framing. Nonetheless, I include them here, along with pain points and hypotheses to make it as explicit as possible the problems I want Cat Herder to solve.


Debra Coordination’s problem statement: Debra is a busy professional and social butterfly who needs to stay on top of appointment information in texts so she can move ahead in her career and maintain important relationships.

Debra’s pain points:

  • Financial: No income if she misses a job interview time and date specified in a text message. Projects will not progress if stakeholders aren’t met with.
  • Product: The search tool is rudimentary and slow. The default organization tools are rudimentary. Old texts get buried quickly by new texts. Finding them is difficult and slow, and then acting on that information without having to reproduce it is also difficult and slow.
  • Process: Using Messages to organize and find old texts is slow and laborious, requiring using rudimentary tools or having to spend time re-entering data in some other app.
  • Support: The nature of the issue has to do with the design of the app itself and the capabilities it offers, so support (which is actually bad in the sense of no in-app tips about how to search) is not the focus of her pain points.

Debra’s hypotheses:

  • Debra needs an app to make it effortless to organize, find, and act upon old texts because so much important appointment information is contained within them, and it’s a difficult and slow process to save it somewhere else. 
  • If Debra could easily organize and find old texts, then she would be more efficient and better at staying on top of appointments.
  • I believe that an app that makes it easy and fast to organize and find old texts will be used by Debra more often and more happily than other means to keep track of appointment information passed over texts.

Joe Reminiscing’s problem statement: Joe Reminiscing is a socially busy person who wants text messages to do more to make him feel connected to the people in his life because he believes good relationships are important and text messages are a crucial way to stay in touch.

Joe’s pain points:

  • Financial: A stretch, but there are ties between feeling good and health, and health is costly to maintain or repair. So emotional wellness can impact your wallet if it goes awry.
  • Product: The search tool is rudimentary and slow. The default organization tools are rudimentary. Old texts get buried quickly by new texts. Finding them is difficult and slow, and then acting on that information (enjoy it or re-share it) without having to reproduce it is also difficult and slow.
  • Process: Using Messages to organize and find old texts is slow and laborious, requiring using rudimentary tools or having to spend time re-entering data in some other app. The search appears to cut off entirely after a certain amount of lapsed time (one interviewee could not find a 10-year-old message). 
  • Support: The nature of the issue has to do with the design of the app itself and the capabilities it offers, so support (which is actually bad in the sense of no in-app tips about how to search) is not the focus of the pain points.

Joe’s hypotheses:

  • Joe needs an app to make it effortless to organize, find, and act upon old texts because so much emotionally rewarding information is contained within them, and it’s a difficult and slow process to find it or save it somewhere else. 
  • If Joe could easily organize and find old texts, then he would be more efficient and better at feeling better and being emotionally connected to people in his life.
  • I believe that an app that makes it easy and fast to organize and find old texts will be used by Joe more often and more happily than other means to reminisce, re-share, and ponder old, emotionally rewarding texts.

User Journey Maps

Because the journey took place inside the Messages texting app, this particular UX artifact may lack the typical thrill of a Hollywood film or Cormac McCarthy story. I also feel like this will change after someone tries a prototype. In any case, it did get me thinking about my own biases and an audio interface for finding and organizing texts. It also inspired me to think of ways to re-surface old texts that had been marked in some way, such as telling the user they have an emotionally rewarding text about a certain location when they enter it.

user journey map for Debra Coordination
Click image to enlarge in new tab.

Click image to enlarge in new tab.

Value Propositions

My primary and secondary research revealed that people have a problem with the lack of a fast and powerful way to organize and find text messages within the Messages app on a phone. Complex and limited solutions exist (Keepster and Groups), but they are serving a narrow business interest, solving a different problem, and are difficult to implement.

Click on image to enlarge in new tab.

Crazy 8s Exercise

I love this activity. It often shifts my perspective on the nature of the problem I’m solving. Sometimes it generates ideas for whole new apps while refining the scope of the one I’m working on. Below is a photo of the (terrible) sketches I made to describe different ways users could organize an individual text message. Further down, I summarize three of the ideas that were most interesting to me.

Click on image to enlarge in new tab.

Convert Saved Text Message to Songs

This idea surprised me, as it seemed like an entirely different way to think about the app, as did the other interesting Crazy 8 results. It, like the others I explore in more detail, goes beyond just saving something. It pushes the user into doing something with the “text worth remembering.” In some ways, that seems closer to the mark—it gets people asking why a text matters to them, perhaps triggering interesting thoughts or emotional journeys that then can be documented. This approach would take the text, throw some music behind it, perhaps convert the text into an automatic singing voice, and prompt the user to sing additional lines.

Put Saved Text Message Into a Story Format Requiring Writing Above and Below It

This idea appealed to me because it takes the memory creation process one step further, going from just noting something significant to encouraging users to react to that information immediately, so to contextualize and inform the emotional experience they’re having as it occurs to them. It would insert the noted text into a document that has writing prompts for paragraphs above and below the text. I’m not sure what those prompts might be. One idea is to emphasize temporality; for example, the paragraph above could discuss what led up to the text of interest or the relationship history behind it. The paragraph below could explore what the user hopes that remembering the text helps them accomplish or feel in the future; what do they hope to get out of saving this text?

Drop Saved Text Message Into User-Designated Folders

This was the most mundane of the ideas, but I liked that it was simple and at least allowed people to solve the main problem of trying to organize their text messages. It also got me thinking beyond creating a new folder system with the app; instead, it would be nice if the app actually just grabbed the text and metadata about it and, for example, dropped it into a new Google Doc in a Google Drive folder the user designated.


Goal Statement

All right, after all of that research and ideation, it’s time to craft a neat sentence of two that describes my solution: a goal statement. Here you go:

Our text-message-enhancing app, Cat Herder [what], will let users organize, find, and re-use texts [what], which will affect people struggling to use their texts to stay organized to make more meaning out of text communication [who] by making it easy, fast, and convenient to label, revisit, remix, and re-share old texts [why]. We will measure effectiveness by the number of app users, the number of personalized folders created by users, and the number of messages that have been organized through the app.

In terms of quantifying success, I’d only track the number of folders and number of messages anonymously. It would also be interesting (but perhaps not indicative of success or failure) to identify the most popular folder names, or the ratio of messages stored in a particular folder to all messages. Or, perhaps if a later feature involves providing auto-sorting rules, much like filters in Gmail, then also tracking the number of these filters users have created in total.


User Flow Diagram

Since Cat Herder operates within the Messages app, I am initially designing it to be accessed when long-pressing on a single text message (which I call STM in the diagram) in a thread, an already-known behavior of Messages users. From there, it’s fairly straight forward to organize the STM or explore other folders the user made in Cat Herder.

Click on image to enlarge in new tab.

Storyboards

I made a fair number of digital films when I was in my 20s, so storyboarding felt very familiar to me. The “big picture” storyboard, featured first, explores the how and why of Cat Herder. The “close up” storyboard delves into the details of the app. Do not hold my mouse pad scribblings of strange beings to any standard, please.


Big Picture Storyboard

Click on image to enlarge in new tab.

Close-Up Storyboard

Click on image to enlarge in new tab.

Wireframe Sketches

In these sketches of the main screen of the Cat Herder app (done using Notability on my iPad Mini), I wanted to include a hamburger navigation icon (which would include links to all of the different ways text can be managed and organized, as well as user settings), a text header, a search bar, and then a number of ways for users to view “cat herded” texts with options to do more. These initial sketches certainly helped figure out what I’d need to present to users and how. Given how rudimentary and similar they were, I include them side by side:

Wireframe sketch one. Click on image to view a larger version in a new tab.
Wireframe sketch two. Click on image to view a larger version in a new tab.
Getting at the Messages interface, where users can checkmark and send messages for saving in Cat Herder. Click on the image for a larger version in a new tab.
A sketch of the screen to manage the metainfo for a single saved collection of texts. Click on the image for a larger version in a new tab.
A more refined sketch of the homepage/dashboard for Cat Herder, which displays an overview of recent activity in the app. Click on the image for a larger version in a new tab.

Robust Wireframes

A joyous journey through Figma brought me to the next stage of the wireframes. The design is simplifying a bit. The basic two screens are: 1) A screen to do detailed management of each saved collection of texts; 2) A screen (the home page or dashboard) to view all the collected texts in different ways; I focus on recency in the example below.

Users of Apple Messages use familiar steps to select the messages they want to save in Cat Herder, then press a button at the bottom to complete the action. Click on the image for a larger version in a new tab.
After sending texts to Cat Herder, users can manage them by adding notes, adjusting tags, favoriting or unfavoriting, programming alerts, or re-sharing the texts. Click on the image for a larger version in a new tab.
The home screen lets users scan a list of their saved texts, organized as they like (recency, favorited, random, alert-enabled, or tag). They click on “Read and Manage” to adjust the details of each saved collection of texts. Click on the image for a larger version in a new tab.
In the tags screen, users default to seeing texts saved under most recently used tag. They can adjust the tag displayed by choosing the tag they want, and can adjust tag settings with the knobs button. Click on the image for a larger version in a new tab.

Lo-Fi Prototype

First you’ll see an image approximating the interactivity of the lo-fi prototype. Below that is an embedded version of the prototype produced in Figma. Neither accounts for every last action in the app, as I didn’t want to get to that level of detail yet.

Click on the image to enlarge in a new tab.

Research Plan

You can view a PDF of my usability study research plan here. It’s a bit odd. I drafted it with a moderated usability study in mind, but found an unmoderated usability study to make more sense (using the service userbrain.com). However, for learning purposes, I left in place the elements of the plan that spoke more to a moderated study.


Usability Study (Unmoderated)

Usability studies are a blast—you really start to see how different people think and use tools. They reveal and test your intentional or unintentional design assumptions or decisions. That giant row of buttons at the bottom of your app? Some users may not think they’re buttons!

I switched to an unmoderated usability study after examining the availability of research participants (busy New Yorkers), my time constraints, and a desire to see how people used the app prototype without me talking them through the process. I explored a number of remote unmoderated usability study services. I first tried Maze. It was impressive and relatively cheap ($33 a month) but users had to perform extremely specific actions. I ended up using Userbrain. It was perfect for somewhat open-ended tests and was more affordable than many other options that NN/g reviewed (which I put into a Google Sheet).

I enlisted five people in the study. They had five tasks and answered a number of questions about their text messaging activities. They also took a System Usability Scale questionnaire, which I modeled after this one from Usability.gov.

Grab a cup of coffee, sit back, and explore the results in this Google spreadsheet.


Affinity Diagram, Patterns, Themes, and Insights

Google suggests grouping your usability study observations into patterns. From there, you develop themes based on the patterns. Finally, themes inform insights, which help designers decide exactly what to change and how. I somewhat merged the theme/insight portion because I immediately jumped to translating observations into actions designers should take.

The affinity diagram is below, followed by my themes and insights.

Affinity diagram for Cat Herder text messaging app usability study
Click on the image to view a larger version in a new tab.

I’ve landed on these themes and insights after reviewing the diagram:

  • Exporting: Users wanted to manage more than one collection of text at a time so the app could function closer to a data exporting mechanism rather than just a fun way to share a few old texts. I need to make multi-select capability.
  • Streamlining: Users wanted to do most if not all actions from a single main listing screen rather than need to go into a single collection to manage it entirely. I need to reconceptualize what information to put in the main screen and detail screen.
  • Filtering: Users needed simpler yet flexible ways to filter the listing of collections in the main pages of the app. I need to consider filtering as a slideout screen, given the nuanced ways I want people to use and navigate the app.
  • Tagging: Users needed a better tagging system, something to clarify how it related to favorites and how to manage individual tags and all tags in groups or as a whole. I need to add a page that is solely for managing tags, rather than hiding tag management inside gear-icon pop-up windows or the detail page of a text collection.
  • Popping: Users needed simpler and more exciting design with clearer indications of what is a button and text labels on buttons to be certain. I need to redesign the buttons and text to be more readable and clear.
  • Confirming: Users need to know that a collection has been sent. I need to add a confirmation screen that users click to return to the app.

Research Presentation – Before and After

I dropped my insights into a Google Slides presentation that describes how I changed the prototype in response to the usability tests. Here it is:


Prototyping A/B Example

I took the designs for another spin, pushing things a bit further to define the buttons and layout more clearly as I played with design decisions around white space, typography, balance, proximity, and so on. Arranging all of the elements clarified how the process of making something real reveals the complexity of the details that make up simple broad concepts. If you find a strong UI designer, hang on to them! Below you can see an example of my adjustments of white space. You can also visit an interactive prototype here.

before white space adjustments
Before white space adjustments.
after white space adjustments
After white space adjustments.