´óÏó´«Ã½

We’ve been running a trial of an exciting prototype designed and developed by R&D called Sounds Daily, a personal radio station that selects and streams content for you based on your preferences. Audience feedback has been a key part of Sounds Daily, and audience participation has helped us shape the trial and is driving how we’re thinking about the results of it.

Over a three-week period, 80 participants explored the innovative world of Sounds Daily on their phones and in-car entertainment systems. This hands-on experience was complemented by in-depth research interviews and dynamic â€˜roundtable’ discussions with 20 of the participants, offering us a rich tapestry of insights. By blending real-world usage with detailed feedback, we were able to both analyse user perceptions and see these insights come to life through the data we gathered, created insights that shine a light on possible progressions on this concept.

We tapped into the daily journeys of regular commuters, focusing in on younger audiences and busy families who haven’t yet embraced ´óÏó´«Ã½ Sounds for their in-car entertainment. This deep dive revealed why ´óÏó´«Ã½ Sounds might not be their go-to choice, offering us a vivid snapshot of their audio preferences and highlighting new opportunities to serve their deeper needs in the crowded landscape of on-demand audio.

We also maximised the opportunity the trial gave us to experiment with quantitative data by:

• Recruiting enough people to make for useful data capture (Standard research study of 6/12 people was not adequate).
• Using the trialists ´óÏó´«Ã½ Sounds usage data from before the trial, to understand the subjects and genres listeners were already interested in.
• Presenting participants with their own usage data, facilitating discussions on how they engaged with the prototype.

As the trial was build on a sandbox of ´óÏó´«Ã½ Sounds - a replica created by the Sandbox labs team to copy the ´óÏó´«Ã½ Sounds platform so that we could run a trial which looked and felt like ´óÏó´«Ã½ Sounds - we collaborated with them to integrate analytics. This meant ensuring that the sandbox was logging all the data typically captured by the ´óÏó´«Ã½ Sounds app, as well as adding in new pieces of analytics specifically related to our users interaction with the prototype.

For the trial we used Tableau to create two dashboards of users behaviour to give us more insights about them during the study. The first showed participants ´óÏó´«Ã½ Sounds usage in the months prior to the trial. This helped us tweak the recommendations users were given to more closely align with their preferences. We also created a dashboard of users interactions with the prototype throughout the trail. This helped us identify users who were having issues or not engaging with the study early on, however, it’s main benefit came towards the end of the trial. We were able to use this dashboard to bring user’s data to them and discuss how they used the prototype. This resulted in much richer answers from users during interviews as we were able to discuss their usage data to jog their memories and help them recall more specific instances of use.

What the participants did

People signed up to 4 weeks of active participation. This included:

• 1 x pre trial questionnaire about their current listening habits.
• 2 x post journey surveys per week, with questions and quick scoring about their most immediate experience of using the prototypes.
• 3 x weekly diary studies, asked for more detail about specific elements within the build, such as the recommendations and AI voice.

By the end of the 4 weeks hand-on trial people were able to give increasingly more informed responses to questions about improvements to Sounds Daily, feelings about personal data and the likelihood of people using Sounds Daily if it was launched on ´óÏó´«Ã½ Sounds.

20 participants engaged with us throughout the trial in a series of in-depth interviews, where we combined structured questions with open-ended discussions. This approach created an engaging, two-way dialogue that not only made the sessions enjoyable but also provided us with invaluable, actionable insights. The clarity and richness of the feedback we gathered were instrumental in guiding our analysis and refining our approach.

One standout activity was 'Curate Your Own Stream', where participants had the opportunity to handpick their preferred content from ´óÏó´«Ã½ Sounds' genres and tags. By the third week of the trial, we used insights from these selections to craft hyper-personalised streams tailored to each participant. In our final interviews, playing back usage data to the interviewees and showcasing these custom streams allowed us to vividly demonstrate how enhanced personalisation can transform the user experience, highlighting the significant impact of targeted content on engagement and satisfaction.

To elevate our research and stay true to the Insights Lab’s vision of immersive, people-centered studies, we hosted intimate round table discussions with participants. These sessions allowed users, who had engaged with Sounds Daily over an extended period, to delve into their specific experiences and needs. By exploring in depth topics—such as the balance between helpful and intrusive personalisation, the role of AI voice in delivering content, and potential coverage of live events—we gained rich collective insights. This approach ensured that our final research report reflected not just individual opinions but also a collective comprehensive view of the strengths and challenges of innovations like â€˜Sounds Daily.’

What we found

Content Recommendations & Discovery

• The podcast world is overwhelming, making it hard for people to find the right content, especially on the move. 
• Good content recommendations require rich metadata, behavioural data, and smart use of personal information.
• People want content recommendations that cater to their wellbeing (e.g. no morning news, relaxing music after work).

Opportunity: A personalised playlist with user-adjustable options could greatly improve content discovery and give a helping hand in understanding what content people want to support their journey. 

Personalisation & User Control

• ‘Personality’ remains the main attraction for ´óÏó´«Ã½ Sounds, whether from hosts, guests, or familiar curators.
• Skipping content doesn’t always mean dislike—it often means "just not right now". 
• Adaptive content could help users avoid missing favourites, and better match their mood or provide balance in shared/group scenarios.
• People want more control within programmes, like skipping specific parts rather than jumping whole sections.
• Users expect some element of surprise from a service that learns their preferences over time.

Opportunity: Tracking skipping and understanding the reasons behind it can greatly enhance insights into user preferences and behaviour.Personalities and niche topics or genres are unique to the ´óÏó´«Ã½, and could provide moments of surprise into the ´óÏó´«Ã½â€™s extensive content catalogue.

Functionality & Experience

• There’s a desire for features like reminders, saving for later, and the ability to resume content.
• There is no single way participants choose between live and on-demand content; the decision is highly contextual and influenced by factors such as who is present, the timing of live broadcasts, and other circumstances. This area requires further exploration.

Opportunity: A valuable next step would be to investigate how live and on-demand content can seamlessly integrate.

By integrating user feedback, behavioural data and contextual cues, Sounds Daily gave us insights into how to move towards a more tailored and engaging listening experience. The Sounds Daily prototype highlighted that the podcast landscape can be overwhelming, and that providing users an easy way to explore Sounds content was needed. It showed that good personalisation and some level of autonomy would be key to providing such a feature. The insights gained from this project will be instrumental in guiding future innovations, refining product strategies to better meet listener needs and as a demonstration of the power of research that goes beyond decision making based on individual use.