How PocketFM is leveraging AI to become the next big thing in audio

Audio company Pocket Entertainment, which produces serialised fiction and other audio series, has ambitious plans, and AI could really supercharge its growth as the company focuses on creating content at scale that few rivals can match.

Backed by Lightspeed and founded in 2018, the Bengaluru company has already logged 100 billion minutes of listening time and is now expanding in a big way across the U.S., the U.K., Latin America and Europe.

Co-founder and CEO Rohan Nayak, who sat down with Fortune India to talk about the company’s rapid global expansion, says AI is at the heart of this push. He also explains why audio series will become the third major long-form entertainment category, alongside movies and TV shows.

Talking about why scale is important for flagship platform Pocket FM, Nayak says the company created a “blockbuster engine” in 2022 to initiate the company’s strategy to scale the audio series category. The company produced hundreds of low-cost pilot episodes and used signals from social media and in-app data to predict what will most likely become hits and what people will pay for.

Explaining the framework behind predicting the next big audio series, Nayak says, similar to fast-fashion retailers that release thousands of designs online to see what resonates, Pocket’s “blockbuster engine” idea came from the way the startup used to source content. “Earlier, we licensed audio-series rights from writers and publishers, similar to how books get adapted into movies. But licensing carried a lot of risk because we couldn’t predict which stories would be hits.”

Just like fashion retailers, PocketFM borrowed that idea for content and created many options quickly to see what users loved, and scaled only the winners. “Because, in the end, the audience knows what works,” says Nayak.

Since audio is much cheaper to produce than video, the company decided to create many pilot episodes instead of committing to full shows. “We began producing about 100 pilots a month, working closely with writers to develop these initial episodes. Our goal was to identify the ingredients of a blockbuster.”

Using tech, PocketFM analysed metrics such as the percentage of users willing to pay for a show and tested pilots on social media to gauge the total addressable market. “By analysing over 10 social-media signals and in-app data, we could predict whether a show might become a hit,” he says.

Pocket’s game-changer idea: Audio content at scale

The PocketFM co-founder claims the company has been able to leverage AI to create audio content at scale. Talking about how the company leveraged AI, Nayak says LLMs aren’t great at AI voice or fiction writing, so the company built its own in-house team in the U.S. and launched two key AI products. The first of them was an AI voice tool for UGC (user-generated content) writers. 

“On the app’s 'Write' section, a writer creates an audio-series episode and, with one click, can choose from about 150 high-quality AI voices fine-tuned for storytelling. The system adds background music and sound effects automatically, and the episode goes live.”

This has been a game-changer for the company as it released a floodgate of users producing content. Previously, too, anyone could upload an audio show, but few did because they had to find voice artists and edit audio themselves. Since launching this product in March 2024, Nayak claims over 50,000 AI-created series have already been produced on the platform.

The second product PocketFM is piloting is an in-house fiction-writing copilot. Nayak says standard LLMs aren’t strong at fiction, and that the company is careful not to replace human creativity. “Those unique, otherworldly ideas only writers can bring. Instead, we fine-tuned large models and built a proprietary tech stack to help writers craft better stories,” he says.

The AI chatbot works much like other models, though it’s specifically fine-tuned to work better with fiction writing. “With the fiction-writing copilot, writers can start with a simple premise—say, “a gamer trapped in a mobile game”—and brainstorm with the copilot to map story arcs, plan episodes, and outline each chapter. When drafting, it suggests openings and endings, helping writers structure episodes and end on strong hooks.”

Nayak said the feature is especially valuable because many of the company’s UGC creators are part-time writers with limited hours and little formal training. “Drawing on our platform’s extensive data, we fine-tune the model without scraping the open web. We’ve also added a language module, cutting the time to launch shows in new languages from 18–24 months to about three,” says Nayak.

Why betting big on audio makes sense

While social media and video streaming have reached maturity, audio streaming, particularly audio series, has emerged as a promising frontier in entertainment innovation. Mirroring the appeal of video series, audio series offers immersive content experiences, is accessible anytime and anywhere, and caters to the diverse entertainment needs of internet consumers.

According to Redseer’s ‘Ear-It-All: The booming world of audio series’ report released in 2024, with around 1.3 billion potential users globally, the audio series market represented a lucrative $21-25 billion opportunity in 2023, which is poised to double by 2027. PocketFM wants to leverage it in a big way.

Nayak believes audio as a content format will continue to grow with less competition, unlike the video format, where the competition is intense. In future, he says, the company aims to double down on audio only, and he is confident it can capture significant market share in it. “Any video product—whether a game or a streaming app—competes with every platform fighting for screen time: Instagram, TikTok, Netflix, Disney, and more. Audio, by contrast, competes only for listeners’ ears.”

He says that due to the overwhelming amount of content, there’s a growing visual fatigue. “Audio is easier to consume: you don’t need to set aside two or three hours as you would for a movie. People can listen during a commute, a morning walk, a workout, household chores, even at work or before bed.” 

Started by three IIT Kharagpur graduates,  Rohan Nayak, Nishanth KS, and Prateek Dixit, Pocket has so far raised around $197 million at a valuation of about $750 million. The company claims to have about 200 million listeners globally, and the platform’s content library is expanding beyond 1 lakh hours. The company is expanding aggressively after the recent Series D funding worth $103 million.

Earlier this month, PocketFM launched AICreator Suite in India, which enables writers to convert stories into fully produced audio series in a short time.

So the ultimate question is, will Pocket succeed in its goal to become the ‘Netflix’ of audio?