Beyond text-to-speech - How SvD turns written articles into exciting audio experiences
2025 Finalist
Overview of this campaign
SvD has invested in audio journalism for years, and found great success with documentary podcasts such as Dynastin and Blenda. However, producing high-quality podcasts is costly and time-consuming. Our goal with this iniative is to find new cost-efficient ways to enlarge our audio story portfolio to attract new subscribers, be accessible a larger part of the day and deliver flexibility in format.
Attract new subscribers: We know from previous podcast documentaries that audio attracts more female and younger subscribers than our regular subscription base. According to Swedish Mediebarometer, approximately 71% of Swedes have access to a paid subscription of streamed music or podcasts. There is a clear potential for SvD to look into paid audio.
Accessible - all day: Audio format can be consumed a larger part of the day than text that requires screen access. We know from user feedback that users ask for the possibility to consume our content in audio-format, to be able to multitask, to get off the screen and to meet reading difficulties or visual impairment etc. According to Swedish Mediebarometer, audio stands for 33% of Swedes’ time spent on different media, compared to 18 % on text. Making our journalism available in audio format makes us more accessible and relevant.
Flexibility in format: The AI revolution will accelerate infinitive versioning of content and immersive experiences, we need to prepare for it. According to Schibsted In/Lab, especially younger users value the possibility to switch between formats, without losing context or nuance. By transforming our text content in different relevant audio formats, we meet an accelerating need for flexibility and enhanced storytelling from new generations of users.
With these objectives in mind, SvD has during 2024 launched an initiative to ramp up audio production in a cost efficient way. While waiting for Scandinavian AI-generated audio to meet standards for paying subscribers, the aim is to go beyond “basic” AI voice generated text-to-speech experiments, to turn written articles into exciting audio experiences that users are willing to pay for.
Results for this campaign
In spring 2024 we decided to pause experiments with audio content created by AI-cloned voice, since we weren't satisfied with the quality. We instead switched focus to narrated articles, where the reporter reads their text in person.
The response was immediately positive, and we now publish 2-4 narrated articles per week. Anyhow, monetization of narrated articles is complicated, since content is already paywalled in text. We turned to our text archive and refined selected old articles with editorial narration and sound design, creating high-quality audio “archive documentaries” at a low cost. Our first “archive documentary”, Giftgåtan – a true crime report – was a huge success. So far, five archive documentaries have been published, with more on the way.
Results:
Subscription sales: Impressive increase in sales from audio format with 71%. We see that audio content attracts subscribers with no previous paying relation to SvD (+75%) and towards female subscribers that we traditionally are underrepresented in our subscriber base (+80%).
Cost efficient production: In Q4 2023 we published on average one produced documentary per month. One year later in Q4 2024 we publish on average 19 audio stories per month, this with just shuffling of resource allocation and almost no investment or resources added.
User value added: Audio stories, in particular narrated articles, have drastically increased in-app listening (+88% during 2024). With a user rating of 4.2 / 5 (over 1.5k users responding) and comments like "Great to be able to listen when you don’t have the opportunity to read. I think SvD has so much good content!" makes us confident that we are doing something right here. We also see that overall engagement correlates, listeners read more articles and are active more days / month than “non- listeners”.
During 2024 we have successfully proven that we can, in a cost effective way transform our text content into high quality audio content, that both adds clear user value for current users as well as attracting new paying subscribers.
