Full Transcript

Dhruv Sharma: Hey there listeners, how is Monday going for you? Today Utsav and I are sitting, this is episode 94 by the way, we're streaming live. Today Utsav and I are sitting with Sriyansh and Darshal, who are the founders of a company called Linktree. Guys, welcome to the show. It's great to have you with us.

Shreyans Sancheti (Linkrunner): It's great to be here, the real time.

Utsav Somani: So guys, so Linktree is a cool name, but before that, let's understand what is MMP. Assume that we're five-year-olds on the show, so explain MMPs to us and then we'll jump into what the company does.

Shreyans Sancheti (Linkrunner): Quick, quick correction, it's Linkrunner. Linktree is a separate company, but yes, MMPs are mobile measurement partners. If you are building an app and you want to scale your app from 100K to 100 million, chances are you're going to be running acquisition or marketing on hundreds of different sources. And it becomes really hard to track which user is coming from which source and what are they doing on your platform. That's where MMPs like Linkrunner come into play, unifying all data across all sources to help customers scale in a ROAS-positive, efficient manner.

Utsav Somani: And I mean, the pitch is basically India's first AI-powered MMP, but you've been doing, I mean, you returned capital at Bluelun as well. What have you learned there that you're now bringing into this journey?

Shreyans Sancheti (Linkrunner): There's so much. I think being a second-time founder is maybe 5% easier than being a first-time founder. It's still really hard. But I think a few learnings is keep the team small and focus on, I think, revenue as the North Star metric. These are the two things that we have been ruthlessly doing at Linkrunner, which we were not focused on during the previous sort of venture.

Dhruv Sharma: Can you guys just open the black box a little bit for us to see and tell us how you actually do the measurement and the attribution?

Shreyans Sancheti (Linkrunner): Yeah, I'd love to start off and then hand over to Darshal because he leads tech. Basically, we connect to the app and all the other sources, be it meta ads, Google, TikTok, Apple, etc. We get data from everywhere and then we show it to the user in a very easy-to-understand format. Darshal, maybe you want to speak about tech?

Darshil Rathod (Linkrunner): Yeah, so we have our SDKs, any platform that you have built your app on, be it Flutter, React Native, Native Android, Native iOS. You put a small piece of code that we have written that helps us collect first-party data, like battery percentage and what you did, when you did, and then we connect to all of our partners like AdNetworks, Google, Meta, TikTok, Snapchat, all of them, and then sort of correlate all of that data to help you understand what your dollar is doing for you in terms of the outcome on the app.

Dhruv Sharma: Do your users sometimes get surprised when you guys get back to them with the numbers on if they're really buying for the buck?

Shreyans Sancheti (Linkrunner): Yeah, in both ways, in fact. No, of course, so both ways. So we recently onboarded a bunch of customers. We worked with people like Blackmail, Bounce, Playo, you know, a bunch of other, hundreds of customers in 10 different countries. And I think the biggest challenge so far has been that Google shows this campaign got 100 installs, Meta showed that this campaign got 115 installs, but it turns out those numbers are not correct. And the real number is the first 100 customers, it was a struggle to get them to believe that, you know, we are correct and that Meta and Google can over-report. And that was an uphill battle, right? We were just trying to change the trust perception of our customers from Google Meta to a link runner or a third party data platform. But by continuously sending them, you know, user level data, CSVs and logs, they were able to change their trust and more into a link runner. So it's difficult, but yeah.

Utsav Somani: That's on the trust side, but there must be some solid engineering challenges to build this for India as well, right? In a post-iOS 14.5 world and Firebase dynamic links, I think there must be something else that you must have done that is your secret sauce.

Darshil Rathod (Linkrunner): Yeah. So fundamentally we have improved in two places and that has allowed us to A, offer a slightly cheaper solution, also more affordable for early stage companies and B, provide real-time attribution data. So that comes in two forms. One, the place where we store our data.

Utsav Somani: Sorry, just to like quantify some numbers, cheaper by a factor of how much?

Darshil Rathod (Linkrunner): It's almost 8x cheaper than any of our competitors. On an average, 6x cheaper. Yeah. For the same amount of data you get.

Utsav Somani: You must be selling this globally as well then? Out of the 40 customers I believe you have on your website, like how many of them are global? I think around 20%.

Shreyans Sancheti (Linkrunner): Yeah, we're at 175 customers. Yeah.

Utsav Somani: Sorry, couldn't get the number.

Darshil Rathod (Linkrunner): Yeah, it was roughly 20% global customers and we cover 10 different countries apart from India. Okay, that's awesome.

Utsav Somani: Yeah. So yeah, 3.8 billion API calls. What's the secret sauce?

Darshil Rathod (Linkrunner): Yeah, so in terms of tech specifically, we did improvements in terms of database. So Clickhouse is our choice of database. It's one of the world's fastest analytical database. Came out like two, three years ago. That's what primarily unlocks a lot of savings that we pass on to our customers. And then again, a lot of core systems that we have built are primarily using AI as a base. So anytime a customer has a question, we are able to turn around and give them a good quality answer based on which they can make their marketing decisions in like a fraction of time with keeping the team extremely lean. So all of those combined sort of give us the edge.

Dhruv Sharma: Guys, I want to go back to the point you were making around trust, right? So when you're trying to match clicks within stores and everyone's coming to the conversation with their own numbers, do you guys, when you point out an attribution number, do you stand by it? Or do you indicate is it a probabilistic number or is it is a deterministic number?

Shreyans Sancheti (Linkrunner): It's deterministic. We only provide deterministic numbers to all of our customers and we give them a confidence score. So, you know, we're at ninety nine percent for almost almost always for all of our installs. And the biggest trust factor that has helped us is that we don't just tell you a number that X clicks or Y installs. We will give you the user level data as well, that this is the user ID. This is the name and the user deals. You can go and check in your internal back end and then match those numbers with us at platforms like Meta. Google cannot and do not provide user level information or details. They just provide the surface level numbers that 10 clicks or 100 installs. And that becomes really hard to sort of really match what the truth is.

Dhruv Sharma: And because you've been paying as an advertiser, I'm assuming sometimes marketers get horrified to to to learn that they've actually been paying for organic traffic all this while within when they didn't have to. OK. And talk a little bit about the integrations also in the products. You explain the the attribution logic with the integrations, including with Martek companies as well.

Shreyans Sancheti (Linkrunner): Yeah, we have a bunch of integrations today. We work with ad networks like Meta, Google, TikTok, Snapchat, Apple search ads, LinkedIn and a bunch of others for other market solutions. We are partners with Netcore, Moengage, CleverTap, WebEngage, RevenueCat, Braze, a bunch of others and all out of the box sort of integration. So if a user wants to get the attribution data, send it to a different platform for a different use case, they can do that seamlessly. We spent, I mean, Darshan and the tech team has spent a lot of time making sure that our developer experience is world class, because one of the biggest turn offs of a new MMP is the developer experience. It just gets stuck over there. And the whole integrations, be it with other platforms, with our platform, we have our own CLI, we have our own MCP. And so it's just, we have cut down integration time from roughly four weeks to maybe four hours on an average.

Utsav Somani: And talk to us a little bit about LinkAI. You can, I mean, now basically use chat like, I mean, like the AI chatbots to basically learn more about such analytics with your product.

Shreyans Sancheti (Linkrunner): Yeah, yeah. The goal is to build an AI CMO, right? There are so many nuanced things that take up so much time of a marketer that AI can handle. And so everything from surfacing specific insights to signals, fatigue, creative analysis, creative velocity, all of that stuff is surfaced by our agent LinkAI. And it also gives actionable insights about twice a day on what our customers should be doing in order to continuously optimize their marketing spend, be it changing creatives or copies or trying out different platforms. And we're still working on it. It's still a work in progress. But as LLMs get better, our product also gets better. So it's sort of like a tailwind.

Utsav Somani: What are your thoughts on agentic marketing? Because you're doing all of this analysis. So maybe you can help them close the loop as well with automating some of this. If you're building an AI CMO, I believe that's your next step as well.

Shreyans Sancheti (Linkrunner): Yeah, exactly. Gershal, do you want to speak a little bit about this?

Darshil Rathod (Linkrunner): Yeah. So right now, through the platform, we provide our customers a set of signals based on which they can act and either reduce the cost per install or help them increase their ROAS. Moving forward, we would sort of want to build a co-pilot where all of these solutions are surfaced on the dashboard. There's an outcome attached to it that you can expect maybe a 30% uplift in ROAS or something like that. And all the market has to do is either approve or deny that decision. If they do approve it, we go into the ad network platform like MetaGoogle and then make the change actually and then track what that change actually did. And then hopefully one day the platform and the models become good enough that all they have to do is give us the monthly budget, ROAS goal that they have, be it with the ROAS or cost per install. And then we do the entire game of getting new users at their targeted CPIs or like helping them achieve their goals probably in terms of marketing. So an autopilot of some sorts, but hopefully starting off with a co-pilot soon.

Dhruv Sharma: I want to take a step back and ask you guys, how has the CMO's life changed because of AI? How has it become simpler? How has it become even more challenging?

Shreyans Sancheti (Linkrunner): Yeah. So a couple of things. I think a CMO's life has become a little bit more challenging. One, because the volume of content that an average human. And so if someone wants to run ads, they need to have a lot more creative velocity. And so for marketing teams and CMOs, they need to produce 10x more content. And that means 10x more campaigns across multiple different sources or ad network. And so that takes a lot more bandwidth to figure out new budgets, new platforms, new creative angles, new content. But I think AI is making that part easier where it's helping with different angles and analysis. But just like how Cursor made coding really simple or Cloud Code is making everyone a developer, there is no such platform for marketing today. And that is the sort of place that we want to be in. How can we enable almost anyone to become a marketer and to run ads on Google and Meta?

Utsav Somani: Talking about Google and Meta, are you recognized by them? Because I believe no Indian MMP is recognized by them yet.

Shreyans Sancheti (Linkrunner): That is correct. It's not just Indian. No Southeast Asian MMP has been recognized by any ad network.

Utsav Somani: What would that recognition bring?

Shreyans Sancheti (Linkrunner): It will bring the global trust to platforms, to a company that is built from India for the globe. Today, there's only about seven recognized attribution platforms by Meta, and all of them are either from US or Europe. No one in this region of the world, which is kind of disappointing, right? That means they don't recognize any platform that has been built from India or Southeast Asia. So we really want to be the first one. We already have partner IDs. We already have the partnership, and we're on track to become a listed partner within the next six to 12 months.

Utsav Somani: Maybe a related question, but when does a startup or a company, at what scale do they approach you? When is the right time to start thinking about attribution?

Shreyans Sancheti (Linkrunner): Probably when you are spending at least maybe a lakh a month on ads or any user acquisition channel, or if you're starting to scale past 100,000 installs per month or 50,000 installs per month. That's when it becomes really difficult to track which user is coming from where and what they're doing on the platform.

Utsav Somani: We today work. How much would they pay you? A customer pay you in terms of getting all of this data?

Shreyans Sancheti (Linkrunner): We charge roughly one rupee per attribute install.

Dhruv Sharma: OK, well, maybe let's ask another interesting and related question. Utsav and I are in competing businesses, right? And let's say Utsav has the near infinite ad budget. I do not. I need to optimize. Can I come to you guys to actually go toe-to-toe with him when it comes to user acquisition? Is that really the proposition?

Shreyans Sancheti (Linkrunner): We take data security and data compliance very seriously. We are fully compliant today with all certifications like GDPR, ISO and SOF2. So we cannot directly tell you what Utsav is doing to get a great user acquisition.

Dhruv Sharma: But that wasn't the question. That wasn't the question. The question was, you know, so constraints breed creativity in a sense. And he's working there with a near infinite ad budget. I have constraints. I have limitations. I really need to stretch the dollar. Is that when I come to folks like you? But the question wasn't doing shady stuff.

Shreyans Sancheti (Linkrunner): Yes. Yes. Our team.

Dhruv Sharma: Yeah.

Shreyans Sancheti (Linkrunner): Yeah. No, 100%. Our teams are more than happy and experienced to tell you what's working, what's not.

Utsav Somani: But there's also one related question that's just come up on our YouTube live. How are you different slash better than AppsFlyer? I think that might help move things along in this direction.

Shreyans Sancheti (Linkrunner): Yeah. Yeah. AppsFlyer is the market leader. They've been alive for 15, 20 years. They have a huge team. But I think we're better than them in three different ways. Number one is our technical integration. It may take four weeks to integrate AppsFlyer or take four hours to integrate LinkRunner. Second is our customer support. We have dedicated account managers and integration specialists to help you in any part of your journey. And number three is our pricing. AppsFlyer charges 6.5 rupees per install. We charge one.

Utsav Somani: Dude, I think this can go really big. Like, what are the limitations right now for you to just scale up from like 40, 50 clients right now to like 500 in the next year?

Shreyans Sancheti (Linkrunner): So we're at 175 customers today. We have scaled significantly. But to go further, we need bandwidth. Right now we're constrained by our team size and bandwidth. We're still trying to get past that. If we had unlimited capital, we would definitely be hiring a team of great engineers to help us build and scale faster. I think.

Utsav Somani: You're happy in terms of gross margins of the product right now or do you want them to get better over time?

Shreyans Sancheti (Linkrunner): They will get better over time. But we're pretty happy with what we have today. We're running on pretty solid gross margins given our technology stack is significantly more efficient compared to any other MMP out there. But, of course, more distributed systems engineers, more infra engineers can help us bring that cost down by even a fourth or a fifth of where we are today. But it's great where we are even today.

Utsav Somani: Dhruv, any final closing questions?

Dhruv Sharma: What's the roadmap looking like, guys? What are the next three, six months? I mean, I don't even know if three or six months is too long a horizon in this world. We want you to give us a window into the future of the company.

Shreyans Sancheti (Linkrunner): You can expect Linkrunner agents to sort of be the cursor for marketing or the cloud code for marketing. And that is what we're aiming to be in the next three, six months. We just want to enable anyone to be able to run ads. So if you're a solo developer who has used cloud code to build their app to get distribution, we will enable you as a developer to run world class ads with great returns on your ad spend. And so I think that enabling anyone to become a 10x marketer is what we're going for in the next three, six months.

Utsav Somani: That's awesome chance. I mean, one final closing question from my side. You've been in this business long enough. You've done a couple of ventures before this. You've returned capital as well. What's one piece of advice that you leave with people who are listening to the show on? How do you shut down your company elegantly? And what do you do when you're about to decide your next step or next move?

Shreyans Sancheti (Linkrunner): I think shutting down Lulon was one of the hardest things that we've had to do. We had a team of 40. We had to lay them off. We had to return capital to investors. But I think the biggest thing that helped us just being transparent and honest with everyone. Right. Why are we doing this and not trying to sugarcoat anything, not trying to hide anything? You know, this just state the facts as it is. Startups have a high rate of failure. So it's not like anyone is not expecting it. Right. It's it's there. And if someone wants to take the risk and join the startup, then they're well aware. And so I think at Lulon, we had given everyone a time, a long period severance. We made sure that we had outplacement services. The entire team was placed within a week in much, much better roles and a salary that they were offered at Lulon. And so. But the thing that I'm most proud about is that I think out of Lulon came, I think, four or five venture backed startups. And I think having a great culture is something that both Darshan and I strive for every day.

Utsav Somani: Amazing. Wish you guys all the best. And thanks, Yash, for bringing on our show today. Thank you. Thank you.

Shreyans Sancheti (Linkrunner): Thank you so much. Thank you so much, Dhruv.

Utsav Somani: Thank you. Cheers. All right, Dhruv, you and me, let's some India positive news today for a Monday. So let's run them down. I think the first one is, I mean, 10 billion in annual profits, Reliance. This is just insane. Their scale is unparalleled. 20,000 retail stores, 500 million mobility subscriptions. They've got 600 million users in three months in this newly created streaming platform called Jio Hotstar. And the Jio IPO is also upcoming. Analysts are projecting 130 to 180 billion in valuations. This will be something that's, I mean, never been tested before on the Indian market. Reliance revenue is 3 percent of India's GDP. So these are significant, significant numbers. But I think what's heartening to see is that this is a company built on oil and now earns more from telecom streaming and retail than from energy. So what a solid story. And they keep on doing these new things. And I mean, disrupting industries. Telecom is something that they've literally single handedly made India one of the largest data consumers of the world.

Dhruv Sharma: Yeah. Yeah, it is quite a story. Absolutely. And it's it's been a tough year for energy anyway. You know, given what's happening in the Middle East and everything from freight to insurance going up to it. It's pretty heartening to see that the consumer businesses, both Jio platforms and Reliance Retail, are now 55 percent, you know, as a share of the group's EBITDA. And there's another milestone in there, by the way, which is so $10 billion in fat and $120 billion in top line revenue. So the first company to, you know, breach that threshold.

Utsav Somani: Wow. Wow. Looking forward to the IPO. All right. Next one. You'll have to run me through this now. Defense Ministry has invited private players, like three of them, to build the next generation of Indian fighter jets.

Dhruv Sharma: What's this about? So I'm assuming you're familiar with the term stealth. It gets used a lot right now. And so, you know, our program, they call it the acronym they've coined is AMCA, Advanced Medium Combat Aircraft. It's going to be our first fifth gen fighter aircraft, which is basically stealth, which means you can't really see its radar cross section is so small. They basically fly without anyone being able to see them. And and so the news in question right now is, you know, the government inviting the private sector. I believe there are bids coming in from Tata Advanced Systems. There's a bid from a consortium with Bharat Forge and Beml. And there's there's a few others. HL is not a part of this bid. You know, again, people are saying different things about why not. But but one of the reasons I've read is because there was a criterion in the tender that said that your order book can't be more than three times your revenue, which is the case for HL because of the because of the Tejas program. So which is to say HL's hands are already full for them to be able to credibly deliver on this AMCA program. And so, yes, you know, two billion dollars are on the line to develop to design and develop the prototype, you know, within this decade. And then 10 times as much money will get unlocked to to actually produce to start producing them at scale. And the other thing I think worth knowing is one of the bidders that started one system is the is the only company in India with a working assembly line right now that's manufacturing aircraft. They did a partnership with Airbus and the C295 is in production. And so they're the ones who've actually done this. The one thing that is not so great about the program to know is that we still have to import the engine. You remember, we've spoken about the power plants, you know, on this on this show before. And so, again, we will be reliant on General Electric G, I think, 414 that that make of engines. But we want to find able to have an indigenous aircraft with indigenous engines and avionics and, you know, fire control systems and everything.

Utsav Somani: Yeah, the numbers that I have is first prototype will be ready in 2030. That's the timeline that they're expecting and they're expecting deliveries in 2035. 120 jets plus will be ordered if trials succeed. And only US, Russia and China currently operate the fifth gen fighters. And but the positive thing is India's defense exports are up 700 percent in just two years, approximately 20,000.

Dhruv Sharma: Because you bring this up, this reminds me that the American and the Russian aircrafts were available to us. The Su-57, the F-35. But then, you know, the thing is, if you're ever if you're importing and if you're not in the indigenizing, the top of the line aircraft is not going to be available to you. So even Israel, even Saudis, like alliance in the US, they have F-35Bs and not A's. They'll never let you have.

Utsav Somani: They'll never let you have the best tech, right? Because, I mean, it's national security. So I think they always want to protect themselves. Like, I mean, if push comes to serve and like there is a war happening, they always want the upper hand. But, I mean, incredible stuff happening around India defense space. I think they've opened up privatization as well. I think the FTA limits have gone up in defense sector also.

Dhruv Sharma: Absolutely. And this will birth an entire ecosystem around it because when three companies can make an aircraft on their own, there'll be so many suppliers and, you know, component manufacturers and so on and so forth. It should be very interesting.

Utsav Somani: Last one, the Bhavya scheme. 33,660 crores for 100 industrial parks. So I think India is now finally getting into the long manufacturing game as well because we need parks of 100 to 1,000 acres. And government has subsidized this because we've got good infrastructure. But I think we need to be competitive in terms of exporting a lot of our manufacturing capabilities. And we've not been able to develop that. India is literally just 1.8% of global merchandise exports right now. And China is 14%.

Dhruv Sharma: And the readout says that these will be like plug and play sort of, you know, industrial parks, you know, and the idea is to bring them up to a stage where they're, you know, investable even on day one. And we were talking about this just earlier, right, how all the pieces are all there. Like you have enough power to go around finally. And, you know, the transport infrastructure, whether it's road, rail, et cetera, is also it's a lot more robust than it was previously. Manufacturing as a share of GDP certainly has to go up from the current 12, 13%. It even slides down sometimes to close to 25%. I believe that is the goal. And so PLIs were step one. This is step two.

Utsav Somani: And they're calling it the China plus one play. And Bhavya is going to be the infrastructure layer under make in India. Our PLI schemes draw electronic exports of 38 billion in FY25. Apple's Foxconn is, of course, getting big. There's Pegatron already in India. So I think all in all, I think positive reaction. Let's hope all of this comes true soon.

Dhruv Sharma: Yeah. Manufacturing and exports. We've been too reliant on services for too long.

Utsav Somani: Yes. Alrighty. That's it from Dhruv and me. And we'll see you on Wednesday at 4 o'clock. Thank you for tuning in, guys. Bye-bye.