Wellness-Monitoring SMS Bot

Background

Care Daily (website) provides AI Caregiver services to help monitor care recipients and to provide a safety net for caregivers/families by having a constant presence in the home. Previous products required the deployment of hardware, which involves greater upfront costs for both the company and the users. As a SaaS company, Care Daily is exploring leaner ways of providing caregiver support services that don't rely on hardware. This project focuses on tracking and improving the well-being of caregivers, senior citizens, and their families.

Research Goal

The goal of this project is to launch an SMS-based bot that asks wellness-related questions to track and chart a user's well-being over time. This will help us 1) identify caregivers who are close to experiencing burnout, and provide additional support to prevent burnout, 2) monitor senior citizen's well-being and suggest medical attention when deemed necessary, and 3) allow adult children of senior users to better track the senior family member's well-being while protecting senior user's autonomy. 

Research Process

I designed a multi-phased, global research plan for the end-to-end development of this product. 

I first identified opportunities in the marketplace with a competitive analysis. Insights were supplemented with concept testing, for which I interviewed potential users using a storyboard. This method was chosen because storyboards easily convey a typical usage situation and allow interview participants to see how a product may be used. The potential users interviewed are caregivers, senior citizens, and family members. 

Once we had a working prototype, I conducted remote, unmoderated usability testing in order to test how easy the bot is to use, and whether the bot's tone of voice is perceived as friendly and supportive. Since the service is not on a separate platform but rather delivered to the user's mobile phone, it made more sense to launch unmoderated (rather than moderated) sessions. I also have plans to conduct a diary study to examine, over the course of a week, how users' experiences are and to make further tweaks before fully launching the product. 

After the launch, I will create an analytics plan to continuously measure user engagement and monitor improvement in wellness. Feedback will be solicited through user surveys to help retain and grow the user base. 

Challenges and Workarounds

Concept Testing
This was my first time interviewing participants. I had three main challenges: recruiting, technology, managing the conversation

The fact that my primary groups of users are caregivers and senior citizens posed a unique challenge in recruiting participants for my research. I wanted to recruit 5 senior citizens, 5 caregivers, and 5 family members for my concept testing interviews, as they represent different groups of users. I tried three different ways to recruit:

 1) I first sent out an email to the internal listserv within Care Daily and asked people to put me in touch with their older family members and anyone who has had caregiving experience.
2) I then posted on Facebook groups dedicated to connecting researchers with caregivers.
3) I even reached out to local senior homes to potentially interview some of their residents in person. 

I had some interest from the Facebook group, and was able to get some family members of Care Daily employees to sign up after multiple solicitations backed by my team and the CEO. I ended up getting fewer respondents than I had originally aimed for, but quickly learned during my interviews that many senior citizens have had caregiving experience themselves, and many caregivers are also senior citizens. This allowed me to get multiple perspectives from the same interview session, which helped me adjust for the lower number of participants. 

Once I connected with potential participants, I was faced with another challenge. Since the interview involved a storyboard depicting a typical usage situation, I had intended to meet with all participants via Zoom so I could easily share my storyboard. However, 3 out of 4 senior citizen participants did not have a computer/Zoom or strongly preferred to speak over the phone. This was an unforeseen circumstance but I managed by describing each step in words instead. Not having a visual aid to rely on allowed me to describe each functionality in even greater detail.  

In graduate school, my training largely focused on quantitative methods. The interviews I conducted for this concept testing were my first opportunity to interview participants as a qualitative researcher. Managing the conversation to stay on topic and ask all the questions I needed to ask was sometimes trickier than I thought. For example, once during an interview with an older senior citizen, they were more interested in telling me their life story than talking about how this bot may fit into their life. During another interview with a former caregiver, they became emotional when talking about their care recipient and spent more time telling me about their relationship than discussing the bot with me. A few times, I tried to steer the conversation back by connecting details from their stories to our bot, but these attempts were not always successful. While these sessions may not have been as efficient as some of my other sessions, I don't necessarily see them as time wasted. If the goal of UX research is to enhance users' lived experiences, then hearing about their lives can add richer context to those lives that we are trying to enhance.