Full Transcript

Dhruv Sharma: All right, let's get this thing started. My opening question today is to Utsav, when we have episodes like the one today, where we just have one set of guests, what do we call these things? Do we call them like diet T-O-N or do we call them T-O-N light or do we call them single scoop?

Utsav Somani: We should find something. We've got two rock stars with us. So let's welcome them on the show. And I mean, they'll educate us about social news, blind, everything like under the sun. So they're doing a lot of things. I think it'll be interesting to unpack this angle with them. So Niket and Harsh, welcome to the show. Thank you.

Harsh Dwivedi (Medial): Thank you so much, Utsav.

Utsav Somani: We're not exactly Shark Tank, but why don't you educate us? What are you doing at Medial?

Harsh Dwivedi (Medial): Sure. So the idea concept for Medial was born out of the idea that, you know, India is a country and a lot of Southeast Asian countries, right? And even the Middle East. Will undergo a startup boom, like just like the US went through. And since there is so much noise on social media nowadays, we would ideally want a platform that is dedicated to serving the startup ecosystem, not just founders, but let's say potential founders, investors, and even employees in the ecosystem, right? People who are not working in traditional MNCs or businesses, but in very high growth startups, right? Their problems, their wants, their needs are completely different. And not every generic platform would be able to, you know, cater to niche audiences, right? So we are very big believers in the idea that social networks will go vertical eventually, either today or 10 years later. But, you know, there'll be smaller sets of people on different social networks, which will be very active and very high frequency. So with that thought, we started Medial.

Utsav Somani: But startup people, typically like even employees, founders, everyone gravitate toward either LinkedIn or Twitter, right? So they spend most of their time there. So what are the needs that your platform is solving very uniquely that these two don't cater to? I'm a big consumer of content on X and LinkedIn. Sometimes I even post as well. So what are the needs? Because you can do jobs there. You can read about companies. And I think on X, people are doing pseudonymous accounts as well, where they can post about just general feedback about startups as well.

Harsh Dwivedi (Medial): Of course. So see, with X and LinkedIn, right, both the platforms present very different kind of challenges. You will see very different audiences, very different kind of discussions, right? Very aggressive discussions in case of X and very formal discussions in case of LinkedIn, right? But the point is that apart from the technical crowd, right, and the technical topics that are being discussed, there's a chance of, of course, very different topics coming up. For example, there is sports, there is politics, and there is a lot of world updates. So there's a chance of people consuming a lot of topics in general, right? But when you are very focused on one specific niche, for example, media, right, what happens is, and this is user feedback from our platform that whether they spend five minutes or 15 minutes on media, and it doesn't matter if they're spending an hour on X or four hours on Instagram Reels, right? What they get to learn is they get to learn something new about the startup ecosystem, right? In general, that is guaranteed non-brain rot, good content that they would consume. And since the discussions are very focused, you will not find any other distractions. So if there's a political post that goes up on media, it immediately gets deleted. So we are very hard that way, right? In terms of moderation, 30% of posts on our platform are immediately moderated. You can't even flinch anywhere, which is not allowed in a topic, which is not allowed on the platform. So in same focus on, you know, making sure that people either learn something or get some kind of benefit.

Dhruv Sharma: Nikit, I often find that people who end up building something akin to a social network have been deeply inspired by something similar in the past. Did you guys have like a starting inspiration? Was it, I don't know, was it HackerRank, StackOverflow, something that you thought was really cool and deserved to be rebuilt for like, you know, for this era?

Harsh Dwivedi (Medial): I mean, of course, like a lot of other kids, our inspiration has been Zuckerberg and Facebook and how the entire social network revolution started. But we were also very obsessed about testing out new and different platforms, right? So I, as a kid, was on MPGH, it was a gaming social forum, right? 4chan, etc, etc. We have seen it all. But we believe that, especially looking at LinkedIn, which is a great platform, by the way, I'm a big fan of LinkedIn too, right? The way they have kept themselves alive and they're still the dominant force in the world is commendable. But yeah, the problems that we noticed being in Mangalore, right? And visiting smaller cities was that the gap was so wide between people who live in, let's say, Bhopal or Lucknow, right? They don't know the startup lingo. They don't know how funds are raised. They just keep hearing about these things on X or LinkedIn and here and there. But unless there is a geographical challenge, right? And that had to be bridged. So it was somewhat social and personal to us that, you know, I shouldn't be the only one since I live in Bangalore knowing about, you know, what liquidity preference is. People should have a bridge to this kind of knowledge and lingo, right? Or in general startup practices. So that was something we were really passionate about that, you know, how do we bring these people on our platform and make tier one and tier two appreciate each other, right? Or learn from each other. Both of them possess different kind of knowledge.

Dhruv Sharma: What have you guys learned about building a social graph underlying the platform?

Harsh Dwivedi (Medial): So if you see our current demographic, and Harsh can correct me if I'm right about this, 49-50% of our audience is from tier one cities, what you can call tier one, right? So Pune, Hyderabad, Delhi, Bangalore, etc. And 50%, almost half of it is tier two, tier three and beyond, right? We are almost present in every city and state. And what we have seen is that having both of these together on the platform and this happened by design, it didn't happen by design, right? It happened just like that. Like we didn't control it per se. We were not putting more ads or more effort in Bangalore, and less effort in some other state, it just happened by chance. And what we have learned about this social graph is very unique, right? The kind of interactions that happen on the platform, and the kind of learning curve that the second part has, right is insane, right? And similarly, it's very unique in nature, like once you use media, you come to know about all these posts that are happening, all these replies that are happening, right? It's very easy, you wouldn't find it on Twitter or LinkedIn.

Niket Raj Dwivedi (Medial): There's no limit in the type of ideas that get based on media, someone might be talking about a Gem AI startup. And someone might be on the other hand talking about a kirana store with some robots flying the deliveries. So very diverse set of people.

Utsav Somani: I mean, most of the platforms like LinkedIn, of course, you add your professional network, but Twitter, or x, as it's called now, I get to choose my own social graph, even Instagram, for that matter, you can follow business pages, you can follow culture pages, you can follow film pages and stuff. So I mean, all of these platforms, including x, I think, is what Neville defines that truly the only platform in the world where you get to curate your own feed, and the algorithm learns and improves with you. So how do you break into the network effects of something like this, where I can curate my own feed, given enough time, like I think it'll learn all my preferences and stuff. So how do you break into that? Like, I think that's the question of breaking network effects to existing social graphs, which are already existing on x, or LinkedIn for that matter.

Niket Raj Dwivedi (Medial): So social has not always been our only one major thing. It has also been that we wanted to cater to everything that is included in the startup ecosystem, like you want to network with founders, and you want to find out investors, we have VC lists, we have angel lists, we have startup news, which covers everyday happenings of the ecosystem. And we also have a launchpad for people to pitch their ideas and get funded right there on the spot. We do three people, three startups a month, we give them grants. So we want to solve all the challenges that a guy who is in any city, wanting to build a startup would do from zero to one, like as funded on a startup of ours called currently, they got like 1 million users after launching a startup showcase. They also came on Shark Tank. And they are doing pretty well, they raised around as well recently.

Harsh Dwivedi (Medial): So just to reframe this, we are not, so the social graph, of course, on x, you know, x catches your interest very well, right? And it shows you the right things that you want to see. Whereas on medial, we do that, we do personalization for a user and the kind of content they want to see on the field, and the kind of news they would like to consume, right? Not everyone will consume, let's say agriculture related startup news, but some people might be more into AI, etc. So personalization, while it is there, our core focus is not just the feed or that consumption, right? There are a lot of features, which are built with the intent of solving the core problems that these people face. For example, the startup showcase, right? It's like a product hunt kind of a forum for launching your ideas and validating them, or just getting new users, finding the right investors, getting spotlight, right? And maybe finding your first few hires, right? So we have certain use cases built in the app that help you all around instead of just maybe you consuming things from that. So you know, we're trying to solve a lot of problems that are very direct to the kind of niche we are targeting.

Utsav Somani: And you talk about niche a lot and vertical, verticalization of social networks. Are there any examples in India? I remember grapevine, which was trying to be the blind of India, there was hood, which was trying to do something related, I think now they're pivoted to dating. So any other vertical social networks that have done well in India? Yes.

Harsh Dwivedi (Medial): So I know of one, which is, I think they've raised more than $7 million. It's some, I think it's called my doc, if I remember correctly. But anyway, there's one app in the pharmaceutical industry, which connects doctors, aspirants, right? People who are studying MBBS, etc, on one single app, and you can actually get mentorship, you can get your questions answered in the medical field, etc, right? And that app is huge, like, the entire medical community appreciates that app a lot. And if you see in hindsight, you cannot get this service on Facebook or LinkedIn or Twitter, right? Of course, you can consume great medical content from doctors all over the world. But the kind of work they have done is very exemplary.

Niket Raj Dwivedi (Medial): Even in the same niche, I think fishbowl and blind have been quite good and fishbowl got acquired by Gloucester. And blind has been scaling.

Dhruv Sharma: And, you know, maybe I'll again ask a follow up to the social graph question. So, you know, when you're literally creating a social graph from scratch, how do hierarchies start to form? Like, do you have people who post more than the others? Who's the power user of the media lab?

Harsh Dwivedi (Medial): Absolutely 100%. So I think 6%-8% of users, of our active users, daily active users, are creating content on a daily basis, right? They create content for sure, every month, and their frequency cannot be matched by a regular poster. It's impossible. So there are users we have seen who have spent four hours a day on the app, and then the average spend time is like eight minutes, right? So compared to them, the gap is huge, the power user gap. So we have definitely seen very weird behavior to answer that question.

Dhruv Sharma: And do you think this could be a way to, I mean, do you guys deep down inside feel this could be a way to equalize things? And what I mean by that is, I mean, in the startup world, capital tends to concentrate, knowledge tends to concentrate, experience tends to concentrate. Is that the overarching goal?

Harsh Dwivedi (Medial): Yeah, the overarching goal is, of course, to create, at least try to create the perfect platform, you know, for the ecosystem, and solve the kind of problems that a normal platform ideally wouldn't even touch, right? So would LinkedIn want to, you know, startup showcase? Probably not, like, because they have to cater to such a varied audience, right? Would they want to provide specific mentorship? Maybe, maybe not, right? Tomorrow, we haven't seen tomorrow. But the idea is that once you focus down, right, then you can just focus on one thing and just eliminate everything else. And that is a superpower, in my opinion, and it will be relevant over the years, whether we will do it or not is not something sure, but somebody someday will definitely, you know, create a large enough social network for techies and people in the ecosystem.

Utsav Somani: Tell us about your algorithm. What goes on inside it? Like, how do you like overweight underweight certain topics? Like, how do you decide that, like, in house? Is there some data science team that you put, or somebody who's driving this engine for you?

Niket Raj Dwivedi (Medial): No, we have like a very simple algorithm right now, because we don't have that many users to like filter the feed for everyone and make it different for everyone. So a lot of because we have less posts, we have to show them to most people. And the basic algorithm that we have is that there is a trending section where every half an hour or 45 minutes, based on interactions and comments, likes, we put posts on the trending section, which is visible to most of the users. And if you are a person with a following graph, or you follow certain people, we have a following feed for you. So it has been very simple. And we have not put much personalization as of now, because that can happen only at a certain scale when we have so many posts.

Utsav Somani: I'll ask a different question. Like, what is the incentive for people to be on your platform and post on your platform? Now, X has started increasing their creative monetization. YouTube, of course, has a big one. LinkedIn is mostly for likes and clicks and brand recognition or personal brand or company brand. So what is the incentive that they get or you have in mind when somebody wants to transition from a user to a power user?

Harsh Dwivedi (Medial): We do have a creator program similar to X, right? But it works differently. So the idea is that the creators who create the most value and not just post the most, whose posts blow up the most, let's say, and garner the most replies, who help each other out, who comment very frequently instead of just posting. So there are a few criteria, 5-6 criteria we have built out. And according to that, we try deciding 10-15 winners every month, up to 30 winners if required, right? So that is there. But I think even today, one of the advantages of being focused is that the moment you hear that a lot of comments and posts will make you bigger, that's not true, right? So what happens is there are no posts which are bad enough for people to get, you know, irked about that. Why is this on the trending feed, right? I wouldn't even want to see it because either it is about AI or it is about dev or it is about creating something valuable, right? So most of the moderation that way is so strict that, you know, we fall into the lane.

Dhruv Sharma: Do you guys know of any platform that allows alumni groups from startups to, you know, stay connected? I mean, when you're working for the company, or if you're on Slack, but you know, if you're a bunch of ex-flipsters, and you'd like to stay in touch with people who've still stayed back, or if you're razors, or, you know, especially some of these large, dense communities, is there something like that? Do you guys have thoughts of that sort?

Niket Raj Dwivedi (Medial): I've heard about this idea, but I've not seen a platform similar to this. Okay.

Harsh Dwivedi (Medial): We actually had a thought of, you know, creating company pages. It's in works. And we had a thought that we'll have a certain tab inside that, that if you were ever part of an organization and you had verified it on media, you will be able to have access to an exclusive feed for people in the company, right? Who are verified. And only people from, let's say, Flipkart will be able to talk on that feed. And those posts will be visible only to the people from Flipkart, let's say. But we haven't executed it yet. It's going to be a bit complicated.

Utsav Somani: I think Grapevine does. And I believe that most of these groups, like what you're mentioning, I think probably are shifting to WhatsApp or private modes of communication. So I think that might be one way that ex-alums stay in touch.

Dhruv Sharma: What's the WhatsApp group limit now? I mean, every time I...

Utsav Somani: It's 1024, I think.

Utsav Somani: It's come a long way. We're going to hit that on the T1 community very soon. So we find out. But let's talk about some fun stuff. Shark Tank, that was a year back. How have the numbers changed? And how has the journey been with Ritesh?

Harsh Dwivedi (Medial): So honestly, our season was probably one of the worst seasons to be on Shark Tank, because it was not on television. It was OTT only. And people weren't really ready to pay Sony Live subscription and watch Shark Tank, right? They would just wait it out for YouTube. So compared to the previous seasons, for example, I know Jasveer, the founder of Hood, and the kind of numbers he had seen, we haven't even seen 10% of that, right? In terms of user bump. So our user acquisition went up by 4 to 5x for a week or two. And our revenue per day went up by 3x for maybe two weeks. After that, it came down. And then again, one month later, when the YouTube episode came out in March, we saw another set of bump, but that was about it. It was more about recognition than actual benefit marketingwise. How many users are there on your platform right now? And you monetize through ads, I'm guessing? No, we monetize through subscription.

Utsav Somani: We have over 500k users. Okay. Interesting. Dhruv, any final closing questions?

Dhruv Sharma: Yeah, I think speaking of subscription, are there parts of the platform that you haven't yet monetized, but you hope to at some point in the future?

Harsh Dwivedi (Medial): So yeah, I mean, 90% of the platform is free. When we started thinking about subscriptions, we introduced two different subscriptions for 99 and 199 rupees. And premium features were built for that, right? So we did not really lock any part of the existing app. We built new features that would be premium and of value to our users. We did research and understood what would people actually pay for, right? And we introduced new features, which we locked behind a paywall. So that is how we had introduced subscription and it seems to be working out for us.

Utsav Somani: And what does the next one year look like for you?

Harsh Dwivedi (Medial): Grow as much as possible, make as much revenue as possible and, you know, probably become one of the least funded but sustainable social networks to have been seen in the country. Awesome, guys. Wishing you all the best on this journey.

Utsav Somani: Thank you. Thank you. Thank you for coming on the show. All right, Ruf, it's just you and me now. Let's run down on the news items. We've got not too much to cover, but I think there's a fascinating report that A16Z puts out every year. Top 100 Gen AI Consumer Apps. The sixth edition came out. Will you take a look at it? I did.

Dhruv Sharma: Who's on top of the leaderboard this time? Is it still the old position?

Utsav Somani: More people in Texas use ChatGP for free than a number of people who pay for Cloud Anthropic. ChatGP is scaling extremely well. 900 million weekly active users up from 500 million in a year, up by 500 million in a year. They are actually 8x of Cloud and 4x of Gemini. So it's still a leader by a mile. India has 100 million weekly active ChatGP users. So that's a number that we'd like to highlight. And A16Z acknowledges that many of these users, of course, are now using these apps and chatbots and these tools via different services as well. So their usage might not actually be reflective of the ranking in this report.

Dhruv Sharma: I think this might just be the third edition of this report. They've been doing it for a while now. And they've...

Utsav Somani: I can't remember. It's the 30th.

Dhruv Sharma: But sixth, for some reason, it tells me. But again, I was saying they've been doing it for a while now. The one thing they've started counting differently this time around is... So earlier, it was just companies that were truly... Like they were built in the AI era. And it's now even those AI-adaptive companies. And Notion, for instance, now is making a lot of... Is adding a lot of new revenue because of their AI features and so on. But even if you look at the... Like right at the top, the duopoly... Or actually, it's not a duopoly yet. The top three contenders are ChatGPT, and Gemini, and Claude. They keep shuffling, yeah. They keep shuffling. The vibes keep shifting. And the models, yeah.

Utsav Somani: And keep on evolving, I think, every three months. Like, I mean, Claude... I mean, I'd never used it until, say, three, four months back in our co-work. And all the features on Claude, I think, are what you rightly said, geared up for a prosumer. Like, I think it's a great graduation from using ChatGPT as a bot that you can chat with and get your queries answered, share some ideas around a doc. But Claude is the one that actually gets stuff done.

Dhruv Sharma: And even the report says, like, ChatGPT is going aggressive on their consumer super app strategy. And Anthropic is not doing that. They want to solve highly specific use cases for you. And so, you know, if you've not been a PowerXL user, they're letting you do that. And then also, as we were chatting just before we started the show, I think they're trying to first erect walls and then bring them down. In other words, you know, create sort of switching costs for people to switch from one to the other. So Anthropic was the first to come up with MCP servers. And people thought this was the way to go. And then now they've productized that in the form of connectors. There's like 50 connectors. And you can get at least six or seven apps to your desktop Claude app.

Utsav Somani: So, yeah, I think web coding, reputable, lovable, Claude code is retentive. Revenue is rising. But top of the funnel growth is also slowing. So the category seems to be maturing. Let's see where this goes. We've got, of course, Emergent, which I hope we'll see on this list very soon, from India.

Utsav Somani: Just to wrap things up today.

Dhruv Sharma: They're calling it agentic engineering now. So by the way, web coding is now... But did you see that report of AWS is calling or some Amazon business called an emergency meeting because of all these outages because of all the web-coded stuff.

Utsav Somani: So... Meta has acquired Moltbook, the agent social network that we discussed. So now, I mean, Meta has one more social network to play with. It's a 60-day old company. The price was undisclosed. They're going to work with the X-scale AI CEO, Alexander Wang. Apparently, Zuck and Alexander Wang were having a beef, but then they clarified that they were not. So, yeah, I think...

Dhruv Sharma: OpenAI and Meta each got one spoil from that war.

Utsav Somani: OpenAI owns the agent runtime. Meta owns the agent social layer. So I think it'll be interesting to see on how they collaborate this thing. Exciting. I think exciting what's happening. And we'll keep you posted whenever we see something interesting. Thank you for tuning in, guys. Bye-bye.