Case Study
Understanding podcaster-fan relationships
During my 12 week summer internship at Spotify, I worked with the Prefab team, which was focused on understanding how we can elevate the podcast-listening experience for listeners and increase user sharing. My internship project served this goal, with the ultimate outcome of highlighting the most promising interactivity concepts to launch. In my work, I collaborated with other user researchers, data scientists, designers, and members of the editorial team​
The research space
The Spotify platform is unique from its counterparts because it hosts both music and podcast content. While both forms of content connect to users through audio, podcasts are different from music in that in many cases, it is difficult to clearly separate the product from its creator. For many podcasts, the product is in fact the host! Consequently, it is not surprising that listeners form parasocial relationships with their favourite podcast creators. This research explores how a new feature on Spotify may support this relationship-building and increase user sharing. Since I cannot disclose the feature( referred to as The Feature henceforth) or describe it, this will be a high level report on how I went about exploring the problem space with my team.
Literature review
Competitor overview
Concept testing
Mapping out the space
Literature review, Competitive overview
I started exploring the research space by first collating past user research on topic, and synthesising insights to bring my team (and myself) up to date with relevant background. The primary objective of this was to understand what previous work had been done to understand The Feature, and what knowledge gaps existed. Much of my work was pulled from past Spotify research, and I supplemented this with academic research relevant to podcaster-fan relationships. I also collaborated with other researchers across the company to validate my insights. Overall, the literature review encompassed 25 internal documents and 10 external research pieces. I followed this with a competitive overview of 6 relevant competitors, with the primary goal of understanding how their platforms support podcaster-fan relationships. I synthesized the findings using affinity mapping and empathy mapping, and based on this, i identified key opportunities for the team to explore in future research. ​
Concept testing
Research planning, developing concepts, user interviews
The primary objective of this stage of research was to understand how The Feature could encourage users to share content from Spotify, as well as attract new users.
Informed by my literature review and competitive overview, I collaborated with other researchers, data scientists and project managers on my team to identify key research questions explore.
Which concepts drive sharing & why?
What value does each concept provide to users?
What are users' expectations for The Feature?
What kind of podcast content do users share & why?
Developing Concepts
In order to understand which concepts for The Feature would support podcast listener sharing, I led a team session to identify which concepts (developed in a prior workshop) would be important to consider for this round of research. The team objectives were to identify concepts that would provide most impact to sharing, but also were feasible as well as concepts that could inform future developments for the feature. From this initial list of concepts, I collaborated with a designer to develop mockups of six concepts across 5 podcast content type domains (e.g health, finance etc) to discuss with participants in user interviews.
User Interviews
We tested our concepts in 75 minute interviews with 10 key participants. We recruited and ran interviews through dscout, and considered the following in selecting participants:
​
Spotify user?
Podcast fan?
Social media use?
We were interested in people who used social media and had a favourite podcast, ​so we could investigate how they supported their listening experiences through social interactions and what kind of podcast-related content they were consuming and sharing online. In the interviews, I discussed sharing and listening behaviour with participants, and collected their feedback on each of the concepts we had designed and tailored to the genres of their favourite podcasts. Because the primary use of the concept testing was to provide feedback for prioritization for my team, I included rating exercises for each concept that described their likelihood to engage and share each concept and supplemented this with a final ranking exercise after we had discussed each concept.
Analysis
Affinity mapping, prioritisation
Because this research was time-sensitive - we wanted to pick a concept to start developing so we could test quickly - my primary analysis method was affinity mapping, with structured coding. I organized findings according to the research questions they addressed and clustered insights within each question. Because the decision of what concept to pursue would influence the entire team, I thought it was important to include the team in this analysis process and lead a workshop to discuss the findings and categorise them. For
This exercise was followed by a later prioritization exercise where we ultimately decided on the most viable concept to work with. In this prioritization exercise, I presented the overall rankings and ratings by participants and then had team members make their own ranking of the top three concepts, which we discussed and used to inform a final decision.
Reflection
I had a fun time working on this project with my team at Spotify. Three of the key skills I built in this experience were managing stakeholders, learning to include team members in analysis and insights building processes, and communicating insights succinctly. My initial plan with this research was to conduct a follow-up survey to get ranking information from a wider user base about user interest in the surveys; however, since I had to cut my internship short for graduate school, this was not possible.