I generally believe we're going to want some generation opportunity to be a world-class internet company out of India. We decided to push ourselves through some really tricky moments. We went from close to zero in scale to like 200 million in like six months. I would not recommend that to anybody.
Really do it for the love of building. That should be the end goal. The end goal should not be at build so I can make XYZ and building so that I have the opportunity to become the one in the building again. Welcome back to another episode of How to Build the Future.
Today, I'm thrilled to be joined by Adit Palachah, the co-founder and CEO of Zepdo, which delivers groceries to your door in just 10 minutes. There are up to 1.5 million orders a day across 50 cities throughout India, with over 3 billion dollars in GMV and growing 300% a year. And there are a unicorn at a $5 billion valuation. Adit, it's great to see you here today.
Thanks so much for having me, Gary. I appreciate it. So when you were getting started, there were some pretty big players already in the space. Swiggy, Zomato, even Amazon was at it.
How did you approach that? When we started and we were very small, there was no sort of grand strategy on ambition, right again, most of the bottom stuff from users in the very early days. But you're looking at it now, I think the reason why we're able to make our dent and create this scale is because a lot of the people that were in the market in the early days, of course, you live in the end there, we're trying to build it supply chain backward instead of user backward or customer backward, right? So they would build out models that were two R-delivy, four R-delivy, which made sense on paper and like, you know, logistically seem more feasible, economically seem more viable instead of building a retention and consumer first.
The insight was, yes, let's do 10-delivy, but also let's build out the full stack of commerce. We were able to run the logistics, we were able to have a much better sense of selection on quality control and sort of that audacious thinking. What is our position? We're not only going to build a better product for users, but over time and now what's playing out is that if you build a better product for users, you also build a better P&L and a better economics, right?
So if users love the platform, your cat is lower, throughput, you're able to get into your supply chain as much higher, your costs come down and now we're basically seeing economics play out pretty beautifully, better than the two R-and-4 R models because of how sticky the users are. So I think basically that's the dent that we had before licensing, just thinking first principle is the user of customer backward versus a lot of the big guys, the big behemoths that had sophisticated P&Ls and finance teams were thinking supply chain backward. I guess if you win the consumer, you can win everything. Yeah, and the big company sometimes will be the side of that.
You're 22 now, but when you were 17, you had reached out to Jared Friedman, one of our partners here at YC. Tell us about that story. So we were just actually supposed to come here in California and study, but the pandemic yet. So the freshman year that we were supposed to have at Stanford basically got postponed by our, we decided to take a gap year and Michael Fonnay was sitting by really had nothing to do, no sort of fancy internships and Goldman or anything like that lined up.
So we just decided to hack around, work on a project and it just started off as like a WhatsApp group where we were delivering groceries for our neighbors because it was a Nike MediK groceries in the first way of the pandemic. And that's slowly updating as just like a fun project and we were talking to people on how to use about it and just sharing updates. And I think one day we saw posting get 10 minutes with our YC partner and we said, wow, this is YC was sort of this two kids on the other side of the world, it was just sort of a mystical land where we were like, you know, this hobby of building cool stuff, people actually go for a living somewhere in the world. Right.
And so we said, wow, we'd love to talk to a YC partner and Jad was kind of like, agree after we, I think we like flooded the comment section saying that, you know, give us time and Jad gave us the benefit. And yeah, we had like, I think it was an eight minute call. We gave him a quick sense. He's like, hey, there's a company in the US called Insta Carda that's something similar to what you guys are doing.
Maybe this could be a business. Have you ever thought of that? We said, you know, not really. But then he said, no, this could actually be something meaningful.
Have you ever thought about YC? And that's the first time we thought, wow, is this like, I thought this was just like a project that actually could be something meaningful. What was the first demo like? How did that sort of come together?
Yeah, I mean, the first like real interaction with the customer was basically WhatsApp group. There was like an old lady down the road from where we were staying in Mumbai that was really struggling to get groceries and she stayed alone. And so we were just delivering groceries for her. And you know, she would basically say, hey, you went to the grocery store down the road.
Why do you want to go to the virtual store? That was like the first year of interaction with the customer. And we should be able to do multiple stores. And then she started adding friends to the WhatsApp group and some more people kept ordering.
But that was the early days. And then I think the first real product that we built was probably I would say two months after the first delivery or like a month and a half, two months after the first delivery. And at the time it was called Kiana Card, right? And we was a big drop service similar to what Instagram is in the US.
That, you know, prototype probably would have taken like 72 hours to build. I attribute most of the heavy lifting to my co-founder, KV, right? But we built it. We asked the users to move to that app.
And that's how it started. There was no like launch moment for say, it was just like getting feedback from customer to WhatsApp and just cycling around. And that just like slowly iterated into something more and more meaningful day by day. There was no like one like boom, you know, startup and then it would actually sort of happen until we got to YC, I guess.
The initial model for it was Karana Card. And then what's the Karana for some of the folks watching? Sure. Okay.
Basically is the Hindi word for mom and pop. Right? So it's like these small mom and pop stores that exist in India. Obviously in the US you've got more large-class goers and googers.
But in India it's primarily these small mom and pop shops that are going to go see. So that's what a Karana card was like corner shopping delivery. That's what it means. So the other thing that was very important was sort of extreme speed.
So it wasn't, you know, an hour or half an hour. It was 10 minutes. How did that come about? So I think when you speak to most people in the US, they think of 10 minutes as a convenience value proposition, but actually India is very different.
So if you look at India, the overall majority of consumption actually happens within four kilometers from your house, which is very different from the US. Like the frequency of purchase in India is four times more for the city than it is in the US. And so people actually just buying these smart tickets multiple times a week, which is why most retail in India is hyper proximity. Right?
So that's the big box one, my Costco models that work like suburban stalls. People are used to the milk delivery guy coming to the doorstep of the food, so that's going outside the house every morning or the local mom and pop in 200 meters down the road. And so that's the dominant format of consumption and retail. Your 10 minutes was never really like a, you know, let's do 10 minutes for the sake of 10 minutes of the week.
It was more, you know, customers were basically planning on the doorstep and we did a lot of the deliveries for the first few months ourselves. They were basically telling us that, Hey, this is great. You know, it's the COVID, but I'm not sure that this is going to be my food, so that's going to be the morning. He comes, you know, pretty much that.
Why do I wait for three hours for you guys? So you guys start the maintenance thing. It's like, I'm just used to buying so much more frequently, so much more easier. And so that's when we sort of gradually moved shorter and shorter delivery times to have the ability to create that doorstep like experience.
Now, obviously it's scaled the, the insight is that it's not just important because people are used to it, but it's just, it's a necessity because you've got a lot smaller household sizes in India. You've got a low penetration of four wheelers. You've got much more perishables to people buy, people have got lower disposable cash. So they don't actually have the luxury of buying in bulk.
And so when you fit all those consumer insights in and mainly just, again, talking to people on the ground, you start realizing that, Hey, this 10 minute thing is actually critical to do the way the style of purchase people like versus like a nice little convenience value problem. So let's talk came out, but I think it was more just talking to people and like constantly saying, you know, reduce it every time by 30 minutes, by 30 minutes, by 30 minutes, eventually you were like, let's just be at your doorstep. One of the things that YC we often talk about is do things that don't scale. It was just one of those things where you had to go and do those initial deliveries at 10 minutes and realized there's actually a palpable difference in that experience from the customer.
100%. So I think in the middle of the batch, we started getting a lot more pushback from users or basically we're just not seeing the attention, right? And the only days we were living like in an hour. And so people would use the app the next week, three, four percent of people use it again.
And so that's when we started talking to people and lots of things came out like selection, pricing, quality, but this was one of the big ones. And so we said, okay, we have to be at the doorstep. So what KV and I did is we went to one of the stores on the platform and we basically commandeered the store ourselves, right? And we ran it for a short period of time, but we did a pilot, but we said if we could, if we were able to control the customer experience and went, including the delivery time, what sort of the action did we get?
And so we basically, like the shop people for a couple of days gave us a little bit of what KV was delivering and we could basically see that at the customer's doorstep, there's an awkward delight and that was obviously the subjective indicator and then as we sort of got to a couple of hundred orders with that model, we could see that people are actually repeating at like, I was the data was not the sophisticated, we could see the very early signs of, hey, this needs like 10 people that ordered last week, but good chunk of them are still ordering this week, maybe that's a good thing. You did this unsalable thing, you did it yourself, you're not running around and hiring other people to do it, you're directly experiencing it, but as a result, you came to sort of ran it to ground and you have this very intuitive idea about what the product needed to become. For sure. And I think that started with the subject of intuition, just like what a human being studying you verbally, you know, what time it became more objective, which is like, what's the retention of each cohort we're seeing?
Are we seeing that repeatability? And I think the, the, the rigor of testing, whether we had PMF came from the, the, the, the grinding we had to go to YC, YC was probably the most intense speed where we learned all these frameworks very quickly. And so Jada would push us saying, okay, you know, you're getting good for other feedback, but a lot of people really, you know, if that's the case, then you're building something real versus just like a hypothesis. You guys also, you know, pioneered this dark store concept that now a lot of people are using, was that also a necessary consequence of becoming a 10 minute provider, like you needed, control over what that experience would be like.
Absolutely right. Like being able to control fulfillment logistics was critical for us to be able to get that, like, that delivery time in our study for the customer. But more importantly than that, you know, the people think of us as a 10 minute provider here. If you go into the ground and you have sort of looked at more as like a, like a hyper local Amazon, not just like good service ability to the customer, but also great selection, great quality, great pricing across.
And you know, today we're doing not like, you know, 700, 800 products, we're doing almost 45,000 to 50,000 products, 30,000,000, exactly everything from like, you know, oranges and apples, all the way to earphones and hoodies, basically bidding out an internet supermarket chain, right? All the people sort of anchor on the 10 minute piece, though owning the full sector to six is also pretty consequential for all the other axioms of customer experience, right? So it's speed for short, and then there's also quality, there's also selection, there's also price. And so all four of those pillars of customer experience improve dramatically, at least not experience, we're able to decide control the entire stack.
So I think as a result of owning it, we're able to, you know, 50,000 products, we're able to control the entire supply chain on the back end and get better quality of produce for customers. Now at scale, we're able to cut all the efficiencies in the back end and give them better prices, right? And so all of that is at the end of the day, head for the customer. So the customer thinks about your 10 minute deliveries, go out from the door, but now they're basically thinking of as they go to like supermarket chain, or they go to one shop shop.
One of the blessings that we had pretty early on is that, like, and you don't remember this as well, right? Very early in the company's life cycle, we got hit by March 2022 with capital markets, which we see for our flip-flip. And we couldn't raise capital, and we had two competitors that had seven times the cash on balance sheet that we had, right? And so that was like a big forcing function, right?
Like, XQ.I was the philosophy of 20, 20, 20, 23. We had to just create as much efficiency, create as much operating level as we could very quickly. Maybe faster than a lot of other consumer-entire e-commerce companies have, globally, much earlier in our life cycle, because it was existential, right? And in retrospect, it was the best thing that could ever happen to us, because we are now capitalized.
We're able to not only allocate the capital, but we're able to do it more efficiently, get more growth for a lot less investment, right? And so essentially what happened is that we got to this point in March, April, 2024, where we had gone from zero in August 2021 to two and a half years later, we had about a billion in GMV. And the business was actually pretty close back then to EBITDA operating cash for Bikivan. And once we were at that point, we were able to show, hey, 70% of markets are profitable.
The most mature markets are touching 4% to 5% EBITDA, and they don't need to be faster and faster. And we're still able to grow, but why don't we accelerate? If the economics are working, then we should actually be accelerating to the business. Instead of trying to cash our chips in at just a billion of GMV, we're going to opportunity to build the largest, you know, gross CPL in India, which is a 700 billion dollar market.
And so if I'm in that position, we can easily build 20, 20 billion of offline, we should be accelerating today. We should be doubling down. And that's what's basically happening at least. What are some of the bumps that you hit in the road, and even as a 22-year-old's founder, how did you think about holding teams to account?
These are sort of the more difficult things, especially in 50 different cities. Where are their cities where the numbers didn't work out and you had to find any leader? Or did you have to react to different specific things in different cities that were surprising? I would say some of the biggest roadblocks have been people have done in the later.
The most fundamental input for our companies is extrusion excellence. And the most important input to extrusion excellence is high quality talent. And so the biggest mistakes I've made as a consequence of that are the wrong highs in some cases or the wrong people manage from. So in retrospect, the setbacks have been, oh, we could have done, let's say, finance better, marketing better, or cashier.
Or category management better, ops better, if we had had the right person. And those mistakes are very costly, especially in that 20-2023 period, but it was existential fuss. So there were situations where we had not so much cash, we almost effectively could have died in that Silicon Valley bank. If you remember that, we had a lot of cash in that period.
Some of those mistakes could have been very costly. And so I think at a high level, just getting the right folks on board, we made some mistakes initially. And now I think we've learned from them that helped us. Let's go a little deeper into some of the existential moments for a zapped out.
What are some other moments that gave you the deepest lessons into building this company and coming up as a CEO? In the very early days, I think we had a very existential crisis that the YC partners had us through. It was like before we got into this whole first party thing, we were realizing when we did an intellectually honest assessment of the business. In the early days, that, hey, there's no tension here.
And this sort of pick up and top off service, they're not able to do it well. And so we thought we were dead. And there was a point in the early days where we just said, what are we going to do? We've tried everything.
We've killed ourselves. This was before the 10 minute. Before the first party started, so this is before we started comment getting started. And we said, this model is not working out.
We thought it was going to be perfect. We just do this model. It's going to grow. We've got a reality check after a few months of getting a wall.
We said, is that even a model that works? And like pretty much everyone's telling us that he goes to delivery in the world. This has never worked really that well. Other than a few exceptions, India, there was lots of companies that came before us.
There was a graveyard. It's all the feedbacks we were getting from investors and people that seemed to know much more than we did is that you guys are essentially in a dead space. And because of the data, so in the very early days, I was like super dispirited, maybe what I mean, like social giving up. That would have been probably the highest likelihood chance of death that this company had.
And obviously as we became a larger and larger company, there were all sorts of crises. I mean, I think we had a few positions where the border race capital and then a crisis happens, whether it's March 22 and the Ukraine, what happened in that period. It's funny actually the most of our countries are coincided with some sort of crazy crisis happening. And it has actually made it harder.
Like I think right before we raised that, around 2023, which is like a hundred years from hell, almost right? Because it was like the only unicorn down in India for that year immediately prior to that, this little very bad thing happened. So today's that it took like eight months, seven months. And within that, I mean, there were lots of places that we could have died in that were real.
Like there was a period where we had to build a lot of financial controls. Like we had to go through like EY audits that if we didn't make it through them as tightly as we would have wanted, we would have not gotten capital. And we would have compared it as a billion dollars of cash in bank, sex, essentially. And we didn't want to sell.
And so we decided to push ourselves through some really tricky moments there, but there probably plenty more still to happen. I think the really cool part of your story is that you didn't look at sort of how other people were doing it and I sort of accepted that as sort of ground truth. There's an X-maxim now that you can just do things. So I see a lot of you can just do things energy.
What you're talking about here? So a little bit of naive and you know, it's like they say like if you knew how hard it was when you started, you wouldn't have done it. Yeah, so I have called Ram Slepliness. I have to say in a nutshell.
So you've personified it. I really like how you thought about retention. Did it feel like you were just asking questions about that and you're out of asking questions? Why is retention not where it could be?
Like that sort of teased out, well, we could try this. We could try that. Like were there a few other things you tried other than the 10 minute thing? Yeah, we tried like a SaaS model for model bar shops.
We tried a bit like a Bw logistics play. Like is that like a better model? Maybe like what if you just do the only delivery part or what if you only do the software part, right? And we just realized that like you know, Grosin India is, you know, can be much better than it is in general as like a marketer.
Like we realize that everything is so broken. Like whether you go to the back end, you go to the store front, you go to the delivery of the last mile. If you want to build something that really solves the problem end to end for the customer, which is across like speed, quality assortment price, then you just need to suck it up. And only the whole thing.
The push basically came from a lot of the psychological candle, right? And I think that you don't really find a lot of that, you know, candidly in India so far, which is why obviously YC was as valuable as it was because, you know, people were just like, you know, brutally honest with us. They're, hey man, you want to make it. You want to build an Instagram card, you want to build a nice consumer company.
This is what the early days of those guys would have looked like. Obviously they went through their own struggles, but you're not, you're not at a point where you can kind of be clean product market fit. And then having that like really brutally honest feedback loop really helped us. And then we sort of pushed ourselves to say, okay, if you're being super honest ourselves, are we building an exceptional product?
Sure, maybe we're getting a little bit of a pop in the early days because it's COVID. People are not going to walk down, they have to use the app. But can I really say that build something that somebody actually wants long term? And until we had a conviction that we were there, we didn't go on it.
It was just like a great feedback loop and like a lot of tolerance for iteration and pain basically. I talked to me about that moment where you did feel like you got a product market fit. And that's when, you know, maybe you were in only one location and then you said, okay, time for, you know, did you do one additional location and do you do one additional location and you say, all right, like five locations. Like what was that like?
If it's actually the first location, first dark story launches in a neighborhood in my microbando and pretty quickly that neighborhood became bigger for us in the rest of the entire city that we were operating in. And it just sort of ballooned and I remember going there because it was an experiment. And it was like effectively, it was like a cardboard cardboard card out almost of like what a dark sort of like today because now dark sort of like you're probably racking and you know, chillers and freezers and everything all set up properly. But over there, just like a bunch of like stuff on the ground basically, we were just living from there.
But the experience was so bad managed because we were able to, you know, curate a good selection, right? And both of the people do that. We were able to get the deliveries consistently. And so just start picking up a very rapidly and long story short, I mean, I did a couple of deliveries and I was like, wow, people really have a spark in their eye when they're getting like a delivery here versus like what I'm used to, which is, oh, this is bad.
This is bad. The deliveries and they had items and not right. So yeah, I think that was like the aha moment. And that is also when we were, we got lucky to just keep getting like the right mentor at the right point of the journey.
And so that's when, you know, our CZ investors, which is next to us in India basically came and said, Hey, you know, this is your pocket fit probably right? And we said, I mean, you're very short, like a very cynical right? That's when the thing that I time that I place, we met the people that gave us the conviction that, you know, you can actually try this again in the second store, try this again in the third store. And that's when it started to be picking up.
We went from basically close to zero and scale to like 200 million in 100 scale in like six months. And like, I would not recommend that to anybody. Like fire, like people really wanted it. Yeah.
Pulling the product out of you. Like when are you coming to my neighborhood? Yeah. And even like just, I mean, obviously it's 2021 since crazy.
Like when you're from a capital perspective, you've been from like effectively the hundred hundred, twenty five K from YC to like a nine million dollar round to a $3 million round to $3 million round in like four months. And that was obviously crazy back then in hindsight, was probably the right thing because that was like sort of the election point to just go on and on. And then in 2022, we stabilized and said, okay, let's start building business office. I guess given that, like, do you consider Zepdo a success?
No, absolutely not. Like, I mean, not even close to it. I think you've got a couple of decades before we can do this. I say that we won.
I generally believe we've got a one-centred nation opportunity to build like a world-class internet company or a window and sitting in places like this, you get really humbled because you realize that, wow, like, look at the kind of quality of companies that have been built in the US and other parts of the world. And I don't think we're there yet. And I think we can kind of be coming and say that, hey, we're able to hold on to some of the best globally, like Amazon or Doraash or people who are always overestimating what they can do in one year and way underestimating what can happen in 2020. Yeah, 100%, 100%.
So you just once tweeted nothing against work-life balance. In fact, I recommend it to all our competitors. Yeah. It's a stupid idea.
I think about that often. I mean, I think it is possible to have balance, but you know, what you can also do, you can run your work like a really good marathon. Yeah, yeah, for sure. And I think for the people that like work at a company like Zepdo, I'd say that, you know, join your, if you're looking to do the best work of your life.
Right. And like, this is the place that is going to get that out of you. Right. So it is a very execution-focused culture.
Everyone is really deep into the game. When you're just like, can you super-excution-focused cultures, then you're really like accountable for real problems every day. You can actually just become like the best version of yourself. And I think that it's not for everybody, right?
That's sort of mindset and thought process. But it is for like the super ambitious, super capable that really want to get the best out of themselves. And it's not for everybody. You're not going to have to do this, like work so hard for many, many decades.
But like, you definitely need like, I think I believe, potions in your life where you take yourself to the maximum, right? And really achieve a lot. I think Zepdo is sort of that place. It doesn't feel like work, at least for me.
I know that's not a realistic expectation for the today. We've got like 3000 plus corporate employees, right? And mainly ops folks, but like with 3000 people. And it's sort of realistic for me to tell everyone that, hey, you know, love what you do.
It works seven days a week. It's not realistic, right? But at least for KV and I, I mean, it just, it feels like the journey of a lifetime. Like, there's nothing more exciting than this to be spending time on.
So that's, I guess, the thought process. But what I will say though is if you don't really have a mission that excites people, right, then you can't push them to that level to really give them. They're all like, SpaceX is a good example of a company that is just able to get the maximum out of people because it's such an inspiring mission. Obviously, we're not sending rockets to space, we're doing the big group series.
But I think that the mission that I like to try to bring out is that guys, let's build like a truly great internet company in India and it's like, take off the revolution in India that should have been much bigger than it is today. I think we can, I think we can create an X more value for the country than we have so far. I think, I mean, if you look at even like places like China, the dynamism came from the internet companies and then that spread everywhere. Right.
So maybe we can be a small part of that. That'd be great. What are some of the unique advantages and challenges you've experienced building startup, you know, specifically in India? So I think the advantage is that the talent is incredible.
Right. And I initially didn't realize this deeply, but you know, we interviewed like American engineers and I was there, except for recently. The guys that we have are as good, right. And a lot of people underestimate that.
That's why, you know, a lot of great internet startups set up offices in Bangalore, right. And high people there. And so the big advantage is you just have such an incredible talent base and really follow the very competitive. It's still a lot less competitive to get such high quality talent that it would be sitting inside from Cisco.
I'm sure. I think some of the hurdles are as an ecosystem. I think we're very much like post 2001 right now because we had like a big reality check in 2023. And there's still a lot of fear and lack of ambition in general.
And I'm sure you've seen some of that like in the post or one kind of period. I'm sure with people who are less ambitious, right, than they could have been. So we're sort of coming out of that and saying, you know, guys, why we, you know, if we execute well, we should go, we should swing for the big leagues. And so I think the challenge is like whether it's hiring a senior executive, whether it's bringing on board a new board member, I feel sometimes and sort of say a little bit in the minority camp.
But we're saying it's very easy to be at like three, four, big enough GMV and just take it easy, right. And just get a big even and then when I say company go public, you do well, right. But wouldn't it be much better to go for five x that six x that and if the opportunity is there, why not go third. And so I think in the US that mindset is sort of taking three hundred everybody has that year.
Right. It's such a special place because of that. Right. But in a place like India, it's still generating.
I mean, I think you're one of the leading importers of philosophy itself. Yeah, hopefully we've got the talent. We just need to get that mindset. The dynamism is like so.
What else do you have on the plate and sort of in your vision for Zeptho over the next 10, 20 years? I think from our perspective, there's a lot of areas to innovate in like in consumer internet. And we want to get distracted where we just, you know, try multiple verticals. Like we really want to just go deep into the compounding flying that we have a product market fit and say, how do we just keep improving our product market fit?
I believe that BMF is like it's not static. It's dynamic. You can keep improving over time if you keep innovating. So there are a few exciting innovations working on Zeptho Cafe, which is basically first party food delivery that we've now started to build into our dark stores where in like a three and a half thousand square foot facility, we have 250 square foot.
That's allocated for coffee, tea snacks, like if you're in a seven to 11 in Japan, for example, even in the US, you have all that amazing in Japan. Not so good here. Yeah, yeah, yeah. Exactly.
And so that it came from the same inspiration that we built out there and that's been phenomenal. So we've scaled that effectively starting last year and we got from zero to over a hundred thousand dollars per day. And that's the beauty of all these consumer internet businesses, right? When you have like 15, 20 million people opening your app every week and then you add like a new, little new case for them.
You can just fund hundreds of thousands of people just by moving pixels around. That's like a super exciting thing about this business. That's why I love it so much. Sounds like it's the true everything store, but it's not only just the everything store that will come to you tomorrow or maybe later next week.
It's literally in 10 minutes. Yeah, it's like the hyper local everything store for hyper local India. That's the idea. But there's a lot of cool stuff.
I mean, even if you look at our advertising business, right? Like some of the other really successful e-commerce companies, if you've gone from like a 40 million ARR in our revenue last year, with this year, we've just crossed 200 million ARR, right? And that's just a, I mean, obviously phenomenal business. And we're just innovating there, you know, exactly.
We built out like our search stack in house, the entire relevance engine, reading, attribution, campaign management, automated keyword suggestions. I think we built out a really high quality performance ad stack out of India for one of the first players to actually be able to do that. Obviously, we've got to fill the focus of that. But we built it really quickly and it's giving us a lot of value in that business.
So ads is another big focus. Cafes are only focused and so on some other stuff. And we're starting to get into, you know, electronics is a category, general merchandise and battle cosmetics. So we're just getting more and more into as many use cases as we can find.
And all of it is user backward. And so I basically spend time with the team looking at, hey, what are people searching for? That we're not able to give them. And then let's start adding those use cases.
So when people started searching for coffee, we said that's put in Zephyr Cafe. When people started searching for lipstick, we said that's not adding cosmetics. So that's the flag is spinning as a result of that. We're getting better attention and we're getting better PNN state as a result of that.
And it's a lot more to push on that. So very early. But I think we could maybe be talking, hopefully, if we execute, it's a big if, but if we execute, well, a few years from now, we'll be talking about a new form of e-commerce. We've sort of invented out of India.
And I think the Chinese guys were able to invent new forms of e-commerce too. So why can't we? Going through your description of how Zephyr works in India, it actually sounds way more awesome than what Amazon is doing. You know, Amazon tried to do some of these faster delivery type things.
They scaled them back. Like we sort of don't know if we'll ever actually even see them. If anything, it's heartening and pretty exciting for the Indian ecosystem. There is this very powerful energy where you can just do things and it's not.
It's happening on the other side of the world. So we're in this age of intelligence, as Sam Altman calls it, and I'm a deep believer in that. What are you seeing? Zephyr, I look at it as a very nascent internet native company.
And I was actually talking to someone at Microsoft yesterday and he gave me a good framework, which is on the, we're basically the application side of the AI acceleration. And if I have to figure out what applications I need to prioritize, because in the end, all the applications are going to be fully fleshed out either to agents or something else that I don't know. But my job right now as CEO is to say, okay, what are the applications that I need to expedite that I need today that are code of my business? And what are the applications that will probably come out of their own and non-code of my business, but I use them, but I don't code in my business.
So for example, let's say non-code will be accounting, right? Can I make my accounting 10 times more efficient, 10 times more accurate? Eventually 100% that the agent that is that for me, can I make my legal contract work that I do at scale? Can I make that much more efficient?
Absolutely. I can. We have a lot of that stuff that we have to do in a direct. But those are not really core.
And someone is going to solve that for me later and I'll pick up an agent or whatever is much more than I do right now, a year from now, nine months from now. And I'll solve that problem. But today, what I'm focused on on the application side is one of the things that are super core that I need today. And I will build some of the capability in house to be able to apply that as soon as possible to my business as an internet native company.
And so for example, let's say the ads business I was talking about, right? So automated keyword suggestion, for example, or bidding attribution, we were able to get much better return and expand for our clients on the ad side. Like you need to go and brock and gamble and go go go go go go go. We were able to do that because we started training on Lama to try to build out a better, to basically build out the relevance that gave us much better results.
So that had a big impact on bottom line. All, for example, the easiest application that we know is customer support. And now that's obviously playing out, but we said we need this customer support stack much quicker. So we built out a crack team that's again, training on some of the foundation models looking on our data.
Let's say like rotten bananas, for example, three to one and a half million hours per day. So we're not going to be delayed on all the other sort of customer support tickets. I'm basically saying, let me just train this model myself or whatever. Let me use what exists out there and give them my data and build out an application for myself and start automating every ticket.
And so today, you know, greater than 50% of the tickets that are raised on Zepro, not just like a rules based chatbot. I actually saw dynamically by a generative chatbot that's been modified for our exact use cases. So we'll focus on these key things like search, ads, customer support, forecasting, supply chain forecasting, another core part of the business. We build a lot of generative, but a lot of deep machine only work that we've done there.
That's exciting. I think one of the observations is that basically these are not solved problems yet. And that's actually exciting. And or even this idea that you could do the everything store across every category and you can basically snap your fingers and add almost any new category.
All of these things are actually empowered by the technology themselves. And those are coming at a more and more rapid pace, actually. Maybe to wrap things up, I guess my favorite question often is what advice would you give the 17 year old version of yourself? Not too long ago, but it feels like in founder terms, you know, the founders live many, many lifetimes every single year.
So what would you say to that person right now? When I started, I was still a little bit extrinsically motivated, but I didn't really done lies that I'm doing this purely for the level building. And that is sort of the majority journey that I've been on when I think the first year or two years or three years of the company when everything was breaking and we were going to die multiple times. And that one, you just basically have a back check where like, what are you doing this for?
And if you're doing it for money or fame or all this other nonsense, then they're far easier ways to do that. And I joke about it, you know, there's far easier ways to make a buck than to build this crazy contraption. But yeah, so I would basically tell them that 17 year old version or whoever else watching this that like really do it for the level building. That should be the end goal.
The end goal should not be I build so I can make XYZ or I can get ABC out of it. But I'm building so that I have the opportunity to become money and build again. And that's basically the mindset that I'm like, oh, every day I just wake up and I'm like, wow, I'm so excited for the next problem statement. And so it's just great.
Thanks so much for spending time with me today. This is awesome. Yeah, thanks guys. Great talking to you.