Ulfa Octaviani

All projects

Industry

Enhanced Podcasts at BBC R&D

A remote field study of an audio-visual podcast player, testing whether images, transcripts and chapters help or get in the way in everyday listening.

Role
UX research intern (end-to-end study)
Timeline
Apr – Sep 2019
Languages
R
Methods and tools
Thematic analysis, Affinity mapping, Surveys
The switch at the bottom right turns the visual layer on and off. Key recommendation: make it opt-in.

Process

  1. Frame the research questions

    Which enhanced features do listeners value? How does the environment change interaction? Does the visual layer help or hinder listening?

  2. Design a field study, not a lab test

    Traded experimental control for realism by testing at home, at work or study, and while travelling. Driving was ruled out for safety.

  3. Recruit and screen

    Recruited 18–35 year-olds, a group the BBC wanted to grow, who listened to podcasts weekly and commuted by public transport or on foot.

  4. Control the content

    Chose three real 10-minute clips across factual, entertainment and current affairs, so genre didn't skew results.

  5. Run five-stage sessions

    Pre-interview and survey, unguided think-aloud exploration, feature training, a week of field listening with surveys, and a closing interview.

  6. Analyse qualitatively and quantitatively

    Thematic coding and affinity mapping of interviews, plus Friedman and Bonferroni-corrected Wilcoxon tests in R on feature and environment rankings.

  7. Report and recommend

    Presented to stakeholders and co-authored the BBC R&D white paper that informed the project's next steps.

Why this study

BBC R&D was prototyping an "enhanced" podcast player that added images, transcripts, chapters and contributor links on top of the audio. Before investing further, the team needed to understand whether these additions changed the listening experience in real, everyday conditions, not just in a lab.

Two screens of the enhanced podcast player for an episode of More or Less. The first shows the episode image, title, description and a vertical chapter bar with Introduction, Sugar consumption, Train fares, Outdoors Play and Planets. The second, during the Sugar consumption chapter, shows a line chart of free sugar consumption falling from 2001 to 2017, with a quote from a contributor below it.
The enhanced player, from the white paper. Left: the episode overview with chapter navigation. Right: a chart and contributor quote shown alongside the audio during a chapter.

Study design

I designed a remote field study so participants used the prototype where podcast listening actually happens. Each participant listened to one clip in each environment over a week, completing a short survey after each session, then joined a semi-structured video interview about their preferred environment, which features they used, and whether they'd switch from their usual app.

Starting with five minutes of unguided exploration captured genuine first reactions before any training. The training step that followed meant later feedback reflected informed use rather than features people simply hadn't found.

Key findings

12 of 13 recruited participants completed the full study.

Home ranked highest, but out of habit rather than focus. It was simply people's default listening moment, and several only glanced at the visuals when something caught their interest.

Travel produced the deepest engagement. Commuters gave the podcast their full attention because nothing else competed for it, and transcripts helped them follow along in noisy places.

Work and study were the weakest fit. Unlike general market research on podcasts, these participants felt their attention was already committed and mostly disliked interacting with the features while working.

Some people treated it like a webpage. A few read the transcript before pressing play, and some found the transcript, images and chapters overwhelming at first. The interface implied a different mental model from the one listeners brought with them.

75% preferred the enhanced player, but most wanted the enhancements to be optional.

Outcome

The main recommendation was to make the visual layer opt-in: audio-only when attention is elsewhere, with enhanced features available when listeners have room for them. This shaped the department's thinking on future iterations, and the findings were published as BBC R&D White Paper 389, which I co-authored.