MICROSOFT RESEARCH
CONTEXT OF INTERNSHIP
During the summer of 2021, I worked as a UX research intern at Microsoft. I collaborated with the Viva Topics product team in Redmond, USA and the Project Alexandria team at Microsoft Research in Cambridge, UK.
I was a JEM intern (Joint Experiences+Devices and Microsoft Research) intern. This meant that I worked across research and product and was co-mentored by both teams. It was a unique experience to be part of Microsoft’s research and innovation lab while working with a product team. It allowed me to feed cutting-edge research on machine-learning and human computer interaction into product development and identify future research opportunities.
*Due to the confidential nature of this work, here in this portfolio, I will only describe the thought and problem-solving process of my internship.
METHODS
- Participant recruitment
- One-on-one user interview
- Focus group interview
- Co-design workshop
- Prototype testing and feedback
ABOUT THE PRODUCT
The Project Alexandria team at Microsoft Research Cambridge works on a machine-learning technology that automatically extracts business knowledge from unstructured enterprise data. This technology powers Viva Topics, an AI product that organizes resources, information and expertise within an enterprise into topics delivered through apps like SharePoint and Office. By connecting people with knowledge and experts, Viva Topics enables faster innovation and collaboration within a company.
Fig 1. Example of Viva Topics in Teams
INTERNSHIP BRIEF
My task was to validate some new feature ideas that the teams have been ruminating on and explore how these features could be designed. The goal was to propose research directions for the team to investigate further. In a nutshell, I did the following:
Conducted user research in order to understand how the product is being used by users and validate assumptions
Sketched up 4 interactive prototypes on Figma to explore what the design could look like and facilitated user research with these prototypes
Communicated user research insights to various product stakeholders and made feature recommendations while taking into account the product's business goals, machine learning capabilities, and development roadmap
PROCESS
1. UNDERSTANDING THE PRODUCT AND THE PROBLEM
I first built rapport with members from various disciplines from both the research and product teams, eg. software engineers, machine learning engineers, customer success managers, designers and product managers. Talking to these people helped me better understand the product and each team's priorities and main blockers. Specifically, connecting with them helped me answer the following questions:
-
What is the business goal for the product and its strategic role in the larger Viva product suite?
-
What is the capability of the machine learning technology right now, and where will it be in 5 to 10 years?
-
What are the main blockers preventing sales or impeding adoption?
-
What are some complaints and praises the team has picked up from different user profiles?
It was also very helpful to establish connections with internal experts from different teams early on, because through them I was able to get questions answered fast and get hands on fresh data or research insights.
2. USER RESEARCH
Since Viva Topics was a brand new product that had just launched in spring 2021, there wasn't any user research done to shed light on how users have been using the product. Therefore, my original plan was to kick off the project by interviewing users and identifying pain-points. However, since most client enterprises were just rolling out the product, we had a lot of difficulty recruiting participants who have been using Viva Topics enough on a day-to-day basis. To avoid being held back by the difficulty to recruit participants, we decided to start the research phase by hosting co-design workshops with internal and external stakeholders.
2*. CO-DESIGN WORKSHOPS
INTERNAL WORKSHOP
Since both the research team and the product team have been ruminating about the features I was tasked to design, I wanted to bring them together through a co-design workshop to collect their ideas.
By creating a space for both teams to brainstorm and discuss ideas together, it helped to clarify each team's priorities and make sure that everyone was on the same page.
Lastly, since members from different roles have their own contact points with customers and users, it was insightful to collate the feedback everyone has collected anecdotally.
EXTERNAL WORKSHOP
Through the product team's user engagement channel, we were able to recruit around 10 knowledge managers from companies that have adopted Viva Topics to participate in the external workshop. Since knowledge managers are people who are in charge of rolling out and managing Viva Topics in their own companies, they have to advocate for the adoption of the product internally and answer concerns and questions from employees. Since we had trouble finding end users who frequently use Viva Topics in their daily work, these knowledge managers are thus the best proxy we could find to understand end user's needs.
Additionally, since users might take on different roles within the Viva Topics system, such as being a knowledge consumer and knowledge contributor, involving knowledge managers in the workshop helped us understand how the product fits in to their company's knowledge ecosystem, and how different user profiles might use the new features.
During the workshops run on Mural and Powerpoint, we did:
-
Inspiration exercise
-
Crazy 8
-
Affinity mapping
-
Group presentation
-
Focus group discussion
Fig 2.Example of a brainstorming board on Mural
[Source: mural.co]
3. SKETCHING
After collecting ideas from the co-design workshops and analyzing the feedback and discussion data, I created 4 interactive prototypes on Figma to explore the design of the new features.
Fig 3. Example of Viva Topics on Outlook
4. PROTOTYPING AND EVALUATION
CUSTOMER ENGAGEMENT PANEL
I demonstrated the prototypes at a customer engagement panel consisted of around 20 knowledge managers from client companies. I used surveys to collect their feedback in order to standardize the response and facilitate comparison and analysis.
1:1 USER INTERVIEW
Towards the end of my internship, I was finally able to find end users who have had ample experience using our product for the interview. Through conducting a thematic analysis of the interview data, I gained insights on how end users have been using Viva Topics in their flow of work and collected feedback on my prototypes.
5. OUTCOME
A. After incorporating the interview data and workshop feedback, I reiterated the original sketches and proposed 4 prototypes to the product team to show how the new features could be designed.
B. I identified 4 future research directions for the teams to pursue. I broke them down into short-term and long-term goals based on the machine learning capabilities, business priorities and the product development roadmap.
C. I communicated my research findings to various product stakeholders including product managers, designers, engineers and researchers.
THINGS I LEARNED
IDENTIFY AND CONNECT WITH EXPERTS EARLY ON
01
I connected with people from different disciplines in both the product and research team early on. Through these connections, I was able to identify experts who can answer my questions quickly or give me access to data and information. These people also referred me to people or resources that I didn’t know would be helpful. They expanded my focus from just Viva Topics to looking at the product suite as a whole. Although I was a new member on the team, connecting with these experts helped me to quickly get onboarded and become familiar with the product’s priorities and strategies. It also made it much easier for me to invite many senior team members to join my co-design workshop later on in the internship.
PRESENT FINDINGS TO STAKEHOLDERS FREQUENTLY TO STAY IN LINE WITH PRIORITIES
02
As an intern, I was not aware of the whole picture of the research efforts and the product development roadmap: what the teams are thinking about is constantly evolving and priorities always shifting. Luckily, my supervisors set up frequent check-ins and progress updates for me with the product and research teams throughout the internship. These check-ins pushed me to frequently synthesize my research and communicate the insights in a clear and concise manner. By keeping the product team in the loop of my research and frequently soliciting feedback from them, I made sure that my project stayed in line with their priorities and that my eventual recommendations would be of value and make an actual impact on the product.
IMPROVISE AND LEVERAGE UPON EXISTING RESOURCES FOR RESEARCH
03
As a student in design school, I always start my design process by understanding the problem. I usually do it through conducting user research, ie. user interview or ethnographic research, in order to uncover the pain points of the people I am designing for. However, through this internship, I learned that it might not always be possible to follow this process when designing and doing design research in the real world: there would always be constraints like the lack of resources or time pressure. As a result, I learned to improvise and leverage upon existing resources to push forward the research, instead of rigidly following the design process and being held back by constraints. I also learned how to apply different tools to the appropriate situations, for example hosting workshops to engage with experts and use them as proxies to understand the problem. Nevertheless, using proxies to understand the problem also requires me to take note of the assumptions I am making and beware of the biases of these proxies.