The Fitness App To Rule Them All

Background:

This was a 10-week quarter long project for my graduate school course covering Human-Computer Interaction (HCI). The primary goal of our team, aptly called Healthy Human Engineering (H2E), was to conduct research focused on improving personal health app usage for Fitbit, our target app.

Our team of 4, scattered across the United States eastern seaboard, conducted comparisons of four competitor fitness products (two direct competitors and two indirect competitors) using heuristic evaluations to identify potential design and usability flaws. We then crafted a survey for Fitbit Users to capture app feedback data based on the identified areas of improvement from the heuristics evaluation, organized our feedback results, and identified 4 User Segments to represent the different users that could use Fitbit. After this, we conducted observation-interviews with Fitbit users (to cover all 4 User Segments) to extract further information on user usage. Lastly, we finished our research with a presentation comparing our survey results with our captured heuristic evaluations and suggested features for Fitbit to implement to address user feedback.

---

Process:

1) Heuristics Analysis (Fitbit and Oura Ring)

I decided to conduct an indirect competitive analysis on Fitbit and Oura Ring, For this analysis, evaluation was specific to two activities: Tracking a walking activity & tracking sleep. Since this was an indirect comparison, it was a chance to see how two different ecosystems attempt to track the same data. Each use case was quite simple: 1) open the application, 2) locate the area to perform tracking of the activity (sleeping or walking), 3) conduct the activity, and 4) verify the app noted completion of the activity. The evaluation criteria contained 13 questions that cover the 10 heuristic principles covered in Nielsen’s Heuristics for Design. Specifically, this evaluation relied heavily on the findings of Affordance, Signifiers, Navigation, Mapping, Feedback and Aesthetics.

For the Fitbit evaluation, the process was directly experienced on my iPhone device. I had installed the app from the Apple app store, used my existing account, and proceeded to complete the two goal with my heuristic evaluation. For the Oura Ring evaluation, the process was more aligned to an indirect observation. I had initially attempted to download the app onto my iPhone device to see how many features were available for users without the ring, but I soon learned that pairing the ring was a portion of the activation process. Thus, the app as I had it downloaded would not function. With this predicament, I resulted in observing other user reviews through reading articles & observing sources through review videos. This material included screenshots of the app’s UI and walked through similar tasks associated with the goals being researched. All material reviewed for this heuristic evaluation were unsponsored videos, so the thoughts & perspectives came from their own experience and not influenced by the parent company.

2) Deciphering Heuristics Analysis

Below are screenshots of the Heuristics Evaluation tool used:

One of the fundamental differences discovered from this competitive analysis was that Oura is more aimed for autonomous wellness tracking. On the other hand, Fitbit has the capability to perform both autonomous tracking and activity-based tracking. The design of Oura’s companion app is minimalistic and contains a dashboard view with related information of readiness & sleep. One of the benefits of this view is it allows users to view their stats at any given time. This can be a good use case for when users are trying to see qualitative data based on how they are feeling, but this also means that they can’t initiate individual workouts or tracking. Fitbit on the other hand has more of a target to track other physical activities, as was the case with Walking and Sleeping, which could have been triggered by the user during any time. This design puts more control into the hands of the user and allows them to customize when they want the application to track their activities.

Another difference was found in the app design & aesthetics. Oura’s app had a very memorable & beautiful layout across all screens. The use of minimal design, color, and spacing made the app feel like the more premium experience. Additionally, the theme was consistent across every screen and the navigation was very natural to interact with. Specifically, the use of color for readiness & sleep statistics was done very well so that users could tell if the findings was positive or negative for health. Compared to the competition, Fitbit struggled significantly, as their navigation was based on some unnatural mapping. During an earlier evaluation, I was consistently confused with the mapping for their app. Additionally, the number of available signifiers for the icons used within the app was not strong, as there was some confusion on where to start tracking tasks during the initial evaluation.

One situation that both apps failed in was with the share-ability of the information. When viewing collected data associated with a user’s profile or tracked data, neither of the apps had the ability to share results. As an example, I could not find an area in the Fitbit app to share any of my personal health metrics or walking data. This could be a useful feature when needed to share health information with a person’s doctor or medical specialist, especially if a user needed to share relevant information about their body’s sleep performance.

3) Survey Design & Questions

For this specific survey, we wanted to ask questions relating to the following constructs:

App Usage Likelihood, App Navigation, and Preferred Features.

For App Usage Likelihood, we wanted to identify how users use the Fitbit app. Specifically, we wanted to investigate matters such as: a user’s preferred frequency of usage, what a user’s primary goal is when using the app, and an overall answer to if the app allows the user to achieve their fitness goals. We chose to have our answer options to be an odd amount to decrease the number of potential ties. Additionally, we wanted each question that qualified to have an available “other” option to capture anything we hadn’t forecasted. From this, we came up with the three following questions for this construct: 1) how frequently you use the Fitbit app, 2) Does Fitbit allow you to reach your fitness goals, and 3) What is your primary goal when using the Fitbit app.

For App Navigation, we wanted to identify how users reacted with the app’s native content layout. With this construct, we specifically wanted to look into the app’s homepage navigation & the ability for a user to find help when needed. From this, we came up with the two following questions: 1) How easy is it to find what you are looking for in Fitbit, and 2) How easy is it to find help within the app while using Fitbit.

For Preferred Features, we wanted to identify what features users enjoyed while using the app. During our heuristics evaluations we had noted that Fitbit had the ability to track calories, walking/running, and sleep. As such, we wanted to use this survey sample to validate the use of those features. Additionally, we wanted to ask the question of which feature was the most & least preferred from their app experience. This was asked to capture any other features that we had not discovered during our heuristic evaluation. Lastly, as a way to track desired features, we wanted to ask a question to see what features they would want to see in the Fitbit app. From all this, we came up with the following questions: 1) how often do you track calories from food, 2) how often do you track your sleep, 3) how often do you track walking/running activities, 4) What additional feature would you like to see in a future Fitbit app, 5) what is your favorite feature in the Fitbit app & why, and 6) what do you dislike the most in the Fitbit app & why.

Based on the three constructs identified, we were able to construct an 11-question survey and shared the survey via many methods including social media posts, company intranet, & sharing with our peers. Our inclusion criteria were the current users of Fitbit and/or those who have used the Fitbit app in recent past.

4) Survey Feedback and User Segments

We set the target number of questions to be in a range of 10-15 to get as much information as possible considering the time to file the survey, which consisted of open-ended and close-ended questions. We built the 11 question survey in web tool called Qualtrics, and it was completed by 21 participants. Below are the top results of the quantitative data of survey:

For the open-ended qualitative data of the survey: some users said they would like to see additional features such as period notification, pregnancy + breastfeeding confirmation, recording blood sugar and blood pressure, etc. We found that as people's standard of living improves, there is a growing demand for "intelligence" and the Fitness app will likely evolve into a "family doctor" in the future.

As a group, we reviewed the trends of data from our Qualtrics survey, and each came ups with data criteria to help define our 4 user segments:

• The Average User

• The Daily Runner

• The Health Nut

• The Occasional User

I was responsible for drafting the criteria for the "Health Nut" User Segment, which is the manifestation of a user who uses Fitbit to track everything about their health and wellness. This includes calories eaten due to food, steps, sleep cycles, and weight. I had discovered this user segment based on the following data points:

Our survey noted that 42% of people use this app daily and noted 3 people saying that Fitbit “Definitely” helps them reach their fitness goal. Based on this data, I identified this user segment targeting a select few of the typical Fitbit user percentage. Another relevant data find for this user segment lies in the amount of people who track their sleep, walking/running, and calories on a daily basis. On average, ~40% track all 3 on a daily basis. Since we are looking at a small user type, I am assuming that this user segment uses these features frequently. Because I am to assume they are very forward thinking on tracking their own data, I am assuming that they find using the app easily. This is because, when averaging all “easy” options for Fitbit use, over 50% of survey responses state how easy the app is to use.

The motivators captured for this user segment are accurate health information, holistic vision of wellness, and actively staying health. This user segment wants to be able to track everything around their health & wellness. As such, they rely on data being readily available and accurate. Since this user is heavily invested in their wellness, they want se the data & information to show them a holistic vision of their wellness. This information allows them to make every aspect of their live healthy, which is their last motivator. For whatever personal reason, they want to be proactive in their health & wellness journey.

The goals that are included in this user segment are: 1) to have accurate tracking of health & wellness, and 2) to be able to have long-lasting health.

5) Observation-Interviews of Fitbit Users

To add another layer of user research, our team decided to conduct personal observation-interviews to better understand the usability of the Fitbit device and its app. We wanted to conduct these observation-interviews with a focus on identifying users that fit into our defined User Segments. As a team we had 4 scenarios for these observation-interviews: Scenario 1 focused on a user tracking a workout, Scenario 2 focused on a user tracking a running activity, Scenario 3 focused on tracking calories, and Scenario 4 focused on tracking steps.

For each observation interview, we asked the following questions:

· How frequently do you interact with the FitBit app?

· Why do you use the FitBit app?

· How do you typically (state the task you are researching) in the FitBit app?

· Throughout this process, what is the easiest part to you? Why?

· Throughout this process, what is the hardest part for you? Why?

I conducted the observation-interviews for Scenario 3. Scenario 3’s participants were recruited via word of mouth. Participant A is remarkably familiar with the Fitbit product and has been using it for over 3 years, whereas Participant B was new to the Fitbit ecosystem.

For Scenario 3, Participant A thought that the app was, overall, easy to use. Being technically minded but not always having fitness in mind, they use the app to track their activities (including food and water tracking). They use the app frequently enough to benefit from having saved food items, so tracking food items for each meal is easy for their workflow. However, it does depend on Participant A to remember to enter in the information.

Participant B had a fresher experience as they we are trying out wearables (by way of Fitbit) for the first time. They also thought that, overall, it was an easy interface. They did have another app previously to track food calories, so their experience is stemmed from a comparison. During the interview, they had the discovery of the ability to scan barcodes for food items, which was a feature in the app they previously used. I also learned that Participant B is not looking for the full bells and whistles for this task, just the ease to get it done.

For tracking calories, Participant B identified with the Average user segment. While wanting to use the app for tracking health, the most important thing that came to mind was the ease of use, which is one of the main desires for the Average User segment. Because of their need to only want to meticulously track everything going into their body, Participant A was a blend of the Average User segment and Health Nut segment.

6) Outbrief Presentation on Research, User Segments, and presentations

After conducting research in the three different ways, we summated our research findings, including suggestions where Fitbit could enhance, and gave a mock presentation. This presentation covered our research methodology, high level results of our competitor's analysis, showcased 4 personas (mapped to our 4 user segments), key research findings, and recommendations for FitBit.

Below are the Persona profiles created for our final presentation:

Below is the findings & recommendation information conveyed in our presentation: