Kicking off from a fairly small section to now streaming 40 billion minutes, together with processing round 4 petabytes of knowledge each month, Pocket FM claims to have constructed the primary audio-entertainment OTT firm, globally.
As per Pocket FM, its content material suggestion system is getting constructed from scratch with none reference framework within the audio OTT house. Additionally it is constructing generative AI capabilities throughout NLP, text2speech, and picture. The corporate claims that the streaming on the platform doubled to 40 billion minutes with the common day by day listener time exceeding 100+ minutes.
Analytics India Journal reached out to Prateek Dixit, CTO, and Co-founder, of Pocket FM, to know extra in regards to the firm.
AIM: How does Pocket FM distinguish itself from the rivals?
Dixit: Pocket FM software program is completely different from the rivals clearly by way of the content material range that it presents. The music or podcast the opposite platforms present creates fatigue after a time. As a listener, I can’t take heed to music on daily basis and even when I’m listening to music for an hour a day, I can’t take heed to the identical track or the identical podcast. It’s all about content material diversification, after which personalised discovery comes on prime of it.
That is the place Pocket FM excels. We offer an incredible quantity of range to our customers. Be it customers of any age group, we offer content material to all with a layer of personalised discovery on prime of that. So you’ll find content material of your curiosity, and I’ll discover mine, as a substitute of everybody being really useful the identical factor after which hitting fatigue after a degree.
So the content material diversification and the library of contents that we offer, after which that candy layer of personalisation to prime it, distinguishes Pocket FM from different rivals.
AIM: How do you propose to compete with key rivals like Spotify?
Dixit: So, to begin with, I believe Spotify is just not a contest of ours. Their major providing is music, whereas our major providing is long-form audio collection content material. Additionally, our audio collection is a bit distinctive in nature, the place we speak in regards to the stats for the content material that we offer. One among our reveals has caught half a billion performs, and round 20 of our audio reveals have crossed 100 million.
That degree of efficacy and stickiness is manner, manner increased than different music or the platform website — folks construct character affinity round these reveals. And since we’re focusing extra on audio collection, the sheer quantity of extra info provides us extra information factors to leverage round making a personalised consumer expertise. I do know {that a} explicit consumer listens to the present for, let’s say, 50 hours. Now, these 50 hours of content material give me an incredible quantity of knowledge to construct suggestions for the consumer. Now, that type of content material may have each implicit or specific sorts of knowledge. It might be, let’s say, the character you want or a canine or the voice of a specific artist, or it could even be the writer. So we then have a whole lot of spectacular survey information factors as properly. And the quantity of knowledge that we now have positioned proper round a consumer’s engagement time is one thing that we utilise to construct for the inhabitants.
We additionally prefer to suppose that the general quantity of consumer information that’s based mostly on the content material we now have is one thing that nobody has, no less than in India. That, kind of, provides us an edge over our competitors to construct higher distribution.
(Pocket FM’s ML Advice System; credit score: Pocket FM)
AIM: What’s your online business mannequin?
Dixit: We’re mainly a freemium mannequin, so we don’t have any arduous paywalls as such. We let customers eat content material free of charge. The best way we monetise the content material is thru the content material windowing technique. So we put paywalls, say, after an hour of explicit content material. We even have launched a micro transaction-based enterprise mannequin; we began microtransactions again in February 2022. And we reached $25 million in annual recurring income final month. That is 12x progress in income in six months.
Pocket FM is an open platform the place anybody might come and create their very own audio present. Once we speak about competing with a number of different gamers, it’s all about making your self defensible, getting the IPs, and creating the know-how behind driving that firm. So we consider that the content material IP, the group, and the AI runs all the product.
AIM: Can AI take over the roles of voice artists?
Dixit: We’re already utilizing that know-how and it’s ultimately going to develop sooner or later, however we consider it’s not going to take over the music world. The musician or the artist is the guts and soul of all the ecosystem, and AI mustn’t take away years of arduous work. However we consider that at a shallow degree, let’s say, creating very small content material of 1 minute, it may be used. But when we speak in regards to the bigger image, we consider that the group, the precise voice artists, will paved the way.
We now have fashions the place we will clone artists’ voices, however by way of utilising these voices for monetisation is one thing that we’re nonetheless not sure about. I believe and not using a correct monetizing mannequin, we will’t even clone the artists’ voices. If we speak about its bigger use case, it may be used to create music on prime of the artist’s voice, which will likely be a useful instrument just for budding artists, and it could albo be monetised.
With regards to generative AI, I believe we’re performing some analysis there; as a tech and AI pushed firm, we consider that we now have to have these cutting-edge applied sciences. We’re pondering by way of creating the cliffhangers within the script with the assistance of huge language fashions.
