FINAL GROUP PROJECT. SPACEJAM FOR BETTERSPACE.

Zacharie Metcalfe, Zeeshan Rasool, James Fearon and Constance Tubbs.

Honestly, I cannot quite believe that I am writing up this project. It really does seem like just yesterday that we were all embarking on this UX journey with General Assembly.

INTRODUCTION

Meeting the client

This sadly was our last project with General Assembly however, excitingly it was our first live client project. Myself and my team of three others were teamed up with Betterspace for this final 4 week sprint. Betterspace is a wellbeing app and resources marketplace. Betterspace enables corporate organisations to purchase subscriptions for their employees which allows them to create goals across the six categories of wellbeing and complete in-app purchases for products and resources. Betterspace is designed for lasting change. The personalisation of the app allows employees to choose the goods and services right for them using the budget set by the employer on a monthly basis. Although we all worked

So, to begin this epic tale of our project it seems only right to take you back to where it all started for us. Coming up with a name for the group seemed the obvious place to kick off. We were working with Betterspace and as we were opening up our first Figjam board of the project we combined the names to come up with SpaceJam. I mean, what is not to love about Michael Jordan and Looney Tunes right?!

We were given access to a demo version of Betterspace and were able to start exploring the app and all it had to offer. It struck us that the app’s development was coming at such a prominent time with the vast effects that the pandemic has caused on day to day life, especially in the work environment. An aspect that we were particularly interested in was regarding loneliness and how we could possibly create features within the app that could help combat it.

During our initial meeting with clients Alfredo and Jim, we established what elements of the brief were to be prioritised during our sprint. The four major priorities we agreed to focus our time on consisted of ; how to encourage users to fill out their self-assessments, how to make data more actionable, thinking about social connection features and finally, improving the overall UX of the user journey. Although we had our main set goals, they were only going to be actionable if users were regularly engaging with Betterspace as a habitual tool to improve their wellbeing to then access these features. How we would think about trying to implement this as a habit is a task we will go into more detail later in the sprint.

INITIAL DISCOVER PHASE

Following on from our meeting, the first task we carried out was our competitive analysis. As a team we felt that it was important to get a better understanding of the competition that already exists to comprehend both how they operate and explore where opportunity lies for our client to grow into. We looked at five competing companies and carried out feature inventory analysis so we could quickly establish elements we knew needed to be worked on or what we were missing altogether. Our findings related back to what we had set out to prioritise initially, reinforcing that we wanted to explore gamification, community, profiles and social connectivity.

Through completing our analysis, we were able to come up with a series of assumptions that would then be used to frame and focus our topics and questions for our user interviews and surveys.

ASSUMPTIONS

Notifications

• Loss of budget in financial terms encourages users to use the app (fear of losing money)
• Users who have no budget need more incentives to use the app and engage with the pillar system

Budget

• Users are more likely to engage with the app if they sent ‘friendly’ reminders to engage

Gamification

• Users may feel hostile with gamification of mental health

The Pillar System

• Users don’t want to engage with the pillar system — they just want to spend their budgets
• Users would be more likely to engage with the pillar system if the rating and self-assessment process was more seamless and unified

Data Visualisation

• Users want an effective and engaging way to visualise their wellbeing data (scores)

Trust

• Trust can be a barrier to engaging with wellbeing apps

Personalisation

• Users want flows personalised to their intended use of the app

INTERVIEWS

In our initial discussion with Betterspace, our client had planned for us to have access to interview several employees from different companies that use the platform. Unfortunately, for reasons unknown to us, we never given any further details and had to pursue with employees from Betterspace themself for data.

Once we had deciphered our overall assumptions, we set about collecting insights through a series of interviews with four employees from Betterspace whilst also creating a more general survey in which we wanted to explore how potential users interact and engage with wellbeing apps. We collectively created our interview questions before we split the team in two so that we would each be able to engage with members of the Betterspace team whilst the other acted as a scribe. The initial questions that we created were designed for us to try and get a broader understanding of how users engage with any wellbeing apps and their relationship with recording any data on their phone whether that be from exercise to mental health. Leading on from these questions we tried to understand how much trust our users put into apps and whether they might feel differently about the platform being provided by their employer.

Personally, this is a stage of research I really enjoy conducting. Building up a rapport during interviews can give way to rich qualitative data that you cannot guarantee by carrying out alternative research. Every employee gave us their own individual insights into the function of the app which allowed us to build a better picture of what the team hoped to see as the future for the platform.

SURVEYS AND KEY FINDINGS

Whilst we were conducting these interviews, we had also sent out our more general survey that we had again created all together.Unlike our Betterspace employees, we had to design the survey with the assumption that the respondents may have never used the platform and so had to include small overviews of how the app functioned before going on to ask them about recording data and trust. Overall, we received 14 responses from potential users oWe were then able to put together an affinity map to better visualise our user data. Some of the most important insights we began to build was surrounding notifications and gamification. Notifications, something that Betterspace does not currently use, our users said could be implemented if done so in an non-intrusive manner, calm tone and controllable frequency. Regarding gamification, this was something we had to ensure was an understood concept by those completing the survey, two thirds of our users were ambivalent but agreed it would be engaging if used in a subtle manner.

DEFINE

Building our persona

Building a persona is a vital part of the UX process to ensure that you are able to visualise the user in mind. It allows both the team and yourself to collectively empathise with your users to ensure that you design with the best intentions for them.

We went on to build our persona once we had collected all of our preliminary data. We created Izzy, a newbie to London, employed by a company using Betterspace as their wellbeing service. Creating our persona came with its own challenges when trying to decipher exactly who our user was. During our preliminary research we had discovered that women were disproportionately negatively impacted in the workplace, we therefore wanted to represent those statistically likely to be in need of Betterspace’s services more. Additionally, given the current state of the work environment in a post-covid world, we wanted to highlight that Izzy was struggling to create a good work life balance given the regularity of working from home. Once we had established Izzy’s persona, we were able to create her user flow and identify the opportunities that were highlighted to better her experience :

- Simplify onboarding

- Reminders

- Budget notifications

• Personalisation of recommendations

USER JOURNEY

Once we had the user flow under our belt, we needed to move onto defining both our problem and how might we statements (HMW’s). This was a really great exercise for the team to do as it helped us collaborate over creating a unanimous goal that we could all collectively work towards. As a team we took our time to try and define our statements as we wanted to be sure which direction we were heading in once we had created an end point to our journey. Our problem statement took some refining before we settled on ‘Izzy needs to be incentivised to be more engaged with Betterspace so that she can improve her mental wellbeing’. We chose this problem statement as we agreed that driving engagement with the platform was the most crucial factor to being able to further assist our users with resources and content. We really fleshed out of HMW’s before deciding on ‘how might we use Izzy’s willingness to improve her wellbeing to get her more engaged with Betterspace?’ Our HMW was built from deciding our problem statement and chosen as we felt that we were going to look into how to build habits within our users and their own desire to want to interact with the app. With these complete, we were able to get cracking with the definition stage of the design.

As mentioned earlier, we were really looking towards making Betterspace a more habitual process for our users. Zach and Zee both identified the Hook Model by Nir Eyal as a great foundation to study how to build good habits for our users. The Hook Model specifically lays out a system for driving the engagement we were looking for and thus felt the most appropriate for what we were trying to attempt within Betterspace’s platform, As you will see here, there are four elements to the Hook Model, they consist of the trigger, action, reward and investment. By implementing these four base actions into the platform, we could hope to see an increased level of engagement between our users and the app.

DEVELOPMENT

Starting design

With this in mind, we moved onto my favourite stage of the design phase, Crazy 8’s. For those unfamiliar with the process it consists of setting an allotted amount of time (usually 30 seconds to a minute) to sketch out a design idea, then… repeating the process eight times over. This stage of design is one I always find the most exciting as it allows you to explore and celebrate the variation of the weird and wonderful ideas between you and your team. Our Crazy 8 day was a great day for the team as it really brought to light the creativity we have on the team whilst simultaneously being able to make us all laugh and comprehend some of the ideas brought to the table.

We completed a set of three challenges for notifications, self-assessments and overall visualisation. We chose these three topics on account of the fact we felt that notifications and overall visualisation were going to be the most effective way for us to drive primary engagement. We additionally wanted to explore what we each had considered for encouraging our users to fill out self-assessments as this is a key step ensuring that our users get a more tailored experience with the app, which we would hope would keep them returning to use the platform. As you will see below these are some of the ideas that we came up with for each feature or design.

Notifications

· Desktop integration outside of the app.

· Creating discreet notifications. Wary that it is an app used for work environment

· Slack integration. Very exciting idea, thinking about some apps out there who have become slack integrated and how we can build on that

· Mobile notifications, thinking about the tone and language used.

· Wellbeing check-ins. This was something we were thinking about outside the self-assessment

· The possibility of a potential social element.

· How we were going to start thinking about self-assessment reminders.

· Budget alerts

Self-Assessments

This was something that we wanted to start thinking about particularly in regards to

· How to encourage users to update their self-assessments. And generally draw them back to the app.

· Potential implementation of a gamification influence in data visualisation. This was something I think we all wanted to investigate how to sensitively integrate

· Possibility of making mobile check-in simpler — how you might be able to reply to notification as a response

· How to access and show past goals and pillars.

And finally the

· Integration of chat feature to check-in.

Visualisation

Particularly

· Visualising data — and how we could create empathy with the user.

· Exploring how to display results for all users to understand.

· How to implement both quantitative and qualitative data with results.

· Gamification elements.

· Calendar display. And how we might further develop that within the app

· Notes and language display in coach marks. Particularly how they might come up after the app hasn’t been used for a while

Completing these challenges gave us an overall better idea of our new user flow and the changes we wanted to think about implementing to reach our prioritised goals. It was at this point that we went on to regroup with our client, Alfredo, to discuss our findings and the direction which we were going.

DESIGN STUDIO

The meeting took place over the course of an hour in which we detailed everything we had achieved up until that point. We were particularly excited about pitching the slack integration as this was something that we had discussed with the Betterspace employees which they had shared our enthusiasm for.

We spent the majority of the meeting discussing the interviews that we conducted and the surveys that we had sent out in parallel. We demonstrated how the data helped us to form our persona, Izzy and how that had enabled us to isolate elements of the brief that we felt most necessary to build for our users. At the end of the meeting we explained our proposed trajectory for the rest of the sprint, explaining the three main elements that we had chosen to work on; notifications, self-assessments and the overall visualisation. Alfredo seemed elated with the work that we had gone through and encouraged us to continue with the plan we had laid out to him.

WIREFRAMES

Before I launch into how we completed all our wireframes, I would like to explain that we made virtually every decision in a collaborative manner. However, as you will go on to read, we did split the team in two for us to achieve as much work as possible within our allotted time frame. James and Zee predominantly worked on our wireframes whereas Zach and I worked on building a notification library, both of which we went on to test. Although we were working on different elements, we were in contact at all times to ensure that both parties were notified and content with any changes being made.

So with that being said, we took our surge of enthusiasm from Alfredo and we moved onto building up our proposed screens. We were fortunate enough to have access to Betterspace’s design library which assisted the speed and quality of the frames that we were constructing.

Self-assessment + Visualisation

As mentioned before, the wireframes were primarily done by James and Zee who did a fantastic job at building up from lo-fi to hi-fi. Before James and Zee had started to build, we had discussed as a team how we were going to integrate the previously mentioned Hook Model, with the addition of elements such as notifications as the first trigger action. Although James and Zee began building in lo-fi, having access to our client’s UI meant that our screens very quickly resembled mid-fi. This was great for us as a team to see it coming together so quickly, and drove us on to wanting to complete the project in the highest standard we knew we were capable of.

As you will see in our mid-fi screens below, we altered the check-in process, so that you would begin by rating the current three goals that you had set rather than grading each of your pillar scores. Users would then be given the option to go on to rate all six of their Betterspace pillars of wellbeing. We added an additional screen after this process to allow users to log any additional information so that they would be able to reference if there was something significant in their calendar that might have altered their scores that week.

As you will see in the mid-fi image below, we altered the positioning of the settings icon from the bottom navigation bar to the top of the dashboard. We wanted to give our users the freedom of being able to choose the frequency in which they receive notifications. This was as our surveys had detailed users specifically wanting the freedom of ability to change this when they please.

After completing the low-fi / mid-fi screens we had initially planned to test the screens before moving onto hi-fi. However, due to our time constraints, we made the decision to get feedback from our team leaders on the course rather than testing at this point. We agreed that it was not ideal and that given more time we would have ensured the lower fidelity screens were tested by users before moving into hi-fi. In hindsight, we can thankfully say that we were able to do this without negative repercussions and thankfully only had to make minor alterations to our hi-fi frames after usability testing. I would like to stress that we made this decision after thoroughly deliberating it through with our teaching team rather than made in haste.

With that all being said, I would like to introduce our hi-fi prototype below.

As you can see below, in the same order, from the first notification through to the proposed user dashboard, it seamlessly looked like Betterspace’s work. The visualisation of the dashboard was something that we really deliberated over as a team, we needed it to be intuitive to use in terms of accessing previous data, as well as being understandable in its overall design, no matter how much we wanted to give it an extreme makeover. We chose to go with a bar graph, influenced by the pillar system already in place, and something our users would easily be able to witness change in when scrolling back through their previous scores.

You will also see here the elements of gamification that we implemented into the platform. The idea of users being able to gain ‘streaks’ fulfilled two goals, not only does it encourage users to return to the app, motivated by the idea of not losing it, but additionally adds that spark of engagement for users to enjoy.

Notifications

The grand reveal of the prototype to Zach and I was superb and really encouraged us all to complete our work to the high standard the others had set. Whilst Zee and James were building out the prototype, Zach and I were focusing solely on building a notification library for the app. We really wanted to explore the tone and language of them whilst also comprehending the personalisation of each one to the individual goal set by the user. This was a really great experience building the library, as we were able to quickly see the variation in our copy and meant that we would be able to test out which one was preferable to the user, if at all.

There are six pillars of wellbeing that allow you to set preset goals within Betterspace ; sleep, exercise, nutrition, meaningful activity, stress management and community. You are able to set three goals across any of these pillars during your engagement within the app, in order for goals to feel manageable and attainable. Within these pillars the goals that you are able to select are predetermined. This meant that there were 72 goals overall which Zach and I split up and went about creating notifications for each of them so that when it came to testing, we would allow our users to select any goal they wished, in order for us to try and gauge the most appropriate style of notification.

Our research had shown us that we would want our notifications to be within 60–120 characters to be easily digestible for our users to interact with. I chose to try and write short, neutral notifications without fully addressing the goal set by the users so that it would act as a reminder to the user, but ideally not inciting criticism if the user had not been able to achieve what they had set out to. I wrote notifications attempting wit — where appropriate, however very much aware that there were goals where it would have not been appropriate addressing topics such as stress management or depression.

Once we had compiled our catalogue, we set about testing them. We created a Google questionnaire where we asked 16 participants to elect a Betterspace pillar and corresponding goal which they wanted to improve on. These participants were again made up of a combination of General Assembly members and friends and family. Depending on their choice, we presented them with the appropriate notification and asked them to give us their opinion and preference. We asked them to compare notifications of varying lengths, personalisation — such as being addressed by their name, use of emojis and tone.

Results from testing notifications

The main takeaways we compiled were that notifications are to be :

- Short, precise and to the point.

- Personal.

- Easy to take in.

- Offer a specific call to action.

Although the majority of users wanted the notifications to be personal, there was a very strong aversion to users being addressed directly by their names. Additionally, users did not want to have emotions projected onto them with prompts such as ‘are you feeling anxious?’ Finally, users also expressed a strong aversion to notifications that attempt to direct them to products in a sales-y manner, but are much more likely to engage with notifications that urge them to use up their budget before the end of the month.

All in all, we got super feedback from our users regarding notifications that we were really looking forward to passing over to our client to talk over. Zach and I were very much aware that some of the goals and pillars within the platform require medical knowledge far greater than ours and would need to be vetted by an appropriately qualified team. We however felt very proud about the amount of data we had been able to give the team in regards to how users would want to interact with notifications for them to provide optimal use.

Testing our hi-fi prototype with our notification library

The final step of testing was to test out the check-in flow that James and Zee had built. This was done after building our notification library so that we would be able to implement some of the notifications that we had tested. We built out our test on Maze and ran 10 unmoderated usability tests which we sent out via Slack. Maze is wonderful as it provides heat maps as well as statistics to make any errors easily identifiable. The main elements we wanted to test were ; the trust within the app, being able to check-in in under 2 minutes and being able to swiftly access both the rating system and the notification settings.

As seen above, we got really positive feedback from our testing, our participants felt that in time they would be able to build up trust within the app and highlighted how this would be encouraged with transparency. As you will see from the statistics and heat maps, our users were all able to get through our desired flow with almost 100% success rates and all in under 2 minutes! Some of our key learnings came when we asked our participants what would stop them from engaging with the check in regularly. Their responses consisted of lack of information and a delay in feedback influencing the personalisation of the app. Overall, we were really beyond thrilled with the testing of the app and really drove our confidence when it came down to our presentation to the client.

NEXT STEPS

Obviously, had we had more time there were multiple elements that we would have loved to have started working on. They are listed below here should Betterspace ever like to have us back to flesh them out further (!) :

Identify opportunities to optimise personalisation within Betterspace features, as a strategy to increase overall engagement with the app.

· Geolocation and thinking about how to allow users to take advantage of what they have available around them

· The development of budget reminders so users are directed to a product catalogue rather than specific items.

· With self-assessments we want to start thinking about how we could update the way users input their data into the assessment and how this should start from an empty state rather than being a pre-set score from a previous assessment.

· And finally, thinking how we could integrate Betterspace into Slack.

CLIENT FEEDBACK

Alfredo and Jim were hugely upbeat in their feedback and wanted to go over the data we had collected specifically from our usability testing. They took a particularly keen interest in how we had come up with our notification library and were ready to look at the quantitative and qualitative data in more detail regarding what landed best.

Overall, they were hugely enthusiastic about the end product iterations that we had made to their design, making note of the way the dashboard had been visualised allowing users to access their past data.

TAKEAWAYS FROM THE EXPERIENCE

This was such a great project to be a part of and hugely insightful as to what to expect when going into my next sprint in the UX world! One of the major elements I took away from the experience was about ensuring time boxing of exercises to allow every step of the discovery and design process to be trialled and tested to guarantee that every move made is backed up with the appropriate user data. Having said this, I am immensely proud of the work that was created and especially the environment that we harvested as a team to allow total freedom of expression and opinion.

Although I hope it was obvious enough to see throughout this case study, I cannot stress how much of a joy it was working with this team. Everyone was consistently there for one another, whether that was a design challenge, to make the others laugh or really just to chat pure gibberish. I really cannot vouch for the combined talent of Zach, Zee and James enough, I should only hope that I have the pleasure to work with them all again in the future.